main
135 次代码提交
| 作者 | SHA1 | 备注 | 提交日期 | |
|---|---|---|---|---|
 Ali Khokhar
|
2a676cc6d9 |
Make ProviderModelInfo the sole model-catalog contract (#1222)
## Problem Provider model discovery consumes metadata, but providers and the cache still expose a parallel IDs-only contract. The duplicate contract adds adapters and lets tests bypass capability metadata. ## Changes | Before | After | | --- | --- | | `BaseProvider` exposed `list_model_ids()` plus a metadata adapter. | `BaseProvider` exposes only abstract `list_model_infos()` returning application-owned metadata. | | Ordinary providers parsed IDs and converted them later. | Ordinary providers parse OpenAI-compatible catalogs directly into `ProviderModelInfo` values. | | OpenRouter, Cloudflare, and GitHub Models maintained redundant IDs-only wrappers. | Provider-specific filters and capability metadata have one return path. | | Vertex returned paginated IDs for the base adapter to wrap. | Vertex returns metadata after completing the same paginated discovery flow. | | The runtime cache exposed test-only raw-ID write and prefixed-ID read helpers. | The runtime cache accepts and returns metadata while retaining its production admin-status ID projection. | | Provider tests asserted the parallel IDs-only API. | Provider tests enforce the metadata-only contract and preserve provider-specific discovery behavior. | | The package version was `4.11.6`. | The package version is `4.11.7` with an updated lockfile. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR makes provider metadata the only model-catalog contract. The main changes are: - Makes `list_model_infos()` the abstract provider discovery API. - Migrates provider parsers and implementations to `ProviderModelInfo`. - Removes IDs-only cache and parser helpers. - Updates provider tests, architecture documentation, and package metadata. </details> <h3>Confidence Score: 4/5</h3> The catalog migration is consistent, but the release version must reflect the incompatible API removal. Repository provider and cache call paths use the new metadata shape consistently. Existing external consumers of the removed contracts can fail after a patch upgrade. The repository rules classify incompatible API removals as a major release. pyproject.toml and the matching package entry in uv.lock <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The external-consumer compatibility probe was run against both revisions, confirming the base still supports the legacy provider contract, while head fails with a TypeError due to the missing list\_model\_infos method, and runtime metadata shows head at version 4.11.7, indicating the removal is a patch transition rather than a major change. - An automated test suite completed successfully with 83 tests passing in 3.31 seconds. <a href="https://app.greptile.com/trex/runs/15207030/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/base.py | Replaces the IDs-only provider API with an abstract metadata-only contract. | | src/free_claude_code/providers/model_listing.py | Consolidates OpenAI-compatible parsing into ProviderModelInfo results and removes IDs-only helpers. | | src/free_claude_code/providers/runtime/model_cache.py | Removes raw-ID helpers while retaining metadata storage and the admin ID projection. | | src/free_claude_code/providers/openai_chat/provider.py | Provides the metadata discovery implementation inherited by ordinary OpenAI-compatible providers. | | src/free_claude_code/providers/vertex/client.py | Preserves paginated discovery while returning metadata values. | | pyproject.toml | Uses a patch bump for a release that removes callable and importable contracts. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart LR A[Provider catalog endpoint] --> B[list_model_infos] B --> C[ProviderModelInfo set] C --> D[Provider model discovery] D --> E[ProviderModelCache] E --> F[Metadata-aware catalog] E --> G[Admin ID projection] ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart LR A[Provider catalog endpoint] --> B[list_model_infos] B --> C[ProviderModelInfo set] C --> D[Provider model discovery] D --> E[ProviderModelCache] E --> F[Metadata-aware catalog] E --> G[Admin ID projection] ``` </a> </details> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22ali%2Fprovider-model-info-contract%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22ali%2Fprovider-model-info-contract%22.%0A%0AFix%20the%20following%201%20code%20review%20issue.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%201%0Apyproject.toml%3A7%0A**Breaking%20Contract%20Ships%20as%20Patch**%0A%0AThis%20release%20removes%20%60BaseProvider.list_model_ids%28%29%60%20and%20cache%2Fparser%20methods%20that%20existing%20integrations%20can%20import%20or%20call.%20Such%20consumers%20will%20fail%20with%20%60TypeError%60%2C%20%60AttributeError%60%2C%20or%20%60ImportError%60%20after%20a%20patch%20upgrade%2C%20so%20this%20incompatible%20API%20change%20requires%20a%20major%20version%20bump%20under%20the%20repository's%20versioning%20rules.%0A%0A%60%60%60suggestion%0Aversion%20%3D%20%225.0.0%22%0A%60%60%60%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1222&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (1): Last reviewed commit: ["Make ProviderModelInfo the sole catalog ..."](https://github.com/alishahryar1/free-claude-code/commit/5c543eadc114201a4085d38885a27ac49153ca29) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45869248)</sub> > Greptile also left **1 inline comment** on this PR. **Context used:** - Context used - CLAUDE.md ([source](https://app.greptile.com/alishahryar1/github/Alishahryar1/free-claude-code/-/custom-context?memory=d2fd24d8-0dec-4faf-8ee4-e085e215a2f8)) <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
1a476562fc |
Make provider model discovery the sole catalog owner (#1221)
## Problem Startup model-list I/O was split between a non-enforcing configured-model validator and the real discovery path. Both queried providers and populated the same cache even though synchronous cache warm-up was the validator's only required effect. ## Changes | Before | After | | --- | --- | | Validation and discovery independently resolved providers, queried model lists, and cached results. | `ProviderModelDiscovery` solely owns model-list queries, failure reporting, and cache population. | | Startup ran configured-model validation and then launched discovery. | Startup synchronously warms referenced providers through discovery, then launches the existing missing-provider background pass. | | Absence from a provider catalog produced a non-enforcing missing-model warning. | Provider catalogs remain discovery metadata and provider execution remains authoritative. | | Configured model references retained environment-source metadata for validator diagnostics. | Configured model references retain only routing data. | | Successful startup providers could be represented by two separate subsystems. | Focused tests enforce concurrent warm-up, single successful queries, failed-query eligibility, and warm-before-background ordering. | | The package version was `4.11.5`. | The package version is `4.11.6` with an updated lockfile. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR makes provider discovery the sole owner of model catalogs. The main changes are: - Warms routed provider catalogs before background discovery starts. - Reuses successful warm results while retrying failed providers. - Removes configured-model catalog validation and source metadata. - Moves query-failure reporting into the discovery module. - Updates focused tests, architecture docs, and package version. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. Provider failures remain isolated and eligible for background retry. Successful warm results are not queried again by the missing-provider pass. Lease release and startup cleanup remain protected by existing control flow. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - \`test\_runtime\_warm\_queries\_referenced\_providers\_concurrently\`, \`test\_startup\_discovery\_queries\_each\_successful\_provider\_once\`, \`test\_failed\_startup\_warm\_remains\_eligible\_for\_background\_refresh\`, \`test\_runtime\_warm\_caches\_all\_referenced\_provider\_models\`, and \`test\_runtime\_startup\_warms\_catalog\_before\_background\_refresh\` all passed. - The complete verbose HEAD run, including command, working directory, exit code, test nodes, and summary, is preserved in \`trex-artifacts/provider-discovery-startup-validation.log\` and the paired after artifact. <a href="https://app.greptile.com/trex/runs/15203372/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/runtime/discovery.py | Centralizes catalog queries, failure reporting, referenced-provider warming, and cache population. | | src/free_claude_code/runtime/provider_manager.py | Replaces validation with discovery-based warming under a generation lease. | | src/free_claude_code/runtime/application.py | Warms referenced catalogs before launching the missing-provider background pass. | | src/free_claude_code/config/model_refs.py | Removes validation-only source metadata while preserving deterministic deduplication. | | tests/providers/test_model_discovery.py | Covers concurrent warming, partial failures, retry eligibility, and single successful queries. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant App as ApplicationRuntime participant Manager as ProviderRuntimeManager participant Discovery as ProviderModelDiscovery participant Provider participant Cache as ProviderModelCache App->>Manager: warm_referenced_model_cache() Manager->>Discovery: warm referenced providers par Provider queries Discovery->>Provider: list_model_infos() end Provider-->>Discovery: metadata or failure Discovery->>Cache: cache successful results Discovery-->>Manager: refresh result Manager-->>App: warm complete App->>Manager: start_model_list_refresh() Manager->>Discovery: refresh only missing providers Discovery->>Cache: cache remaining catalogs ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant App as ApplicationRuntime participant Manager as ProviderRuntimeManager participant Discovery as ProviderModelDiscovery participant Provider participant Cache as ProviderModelCache App->>Manager: warm_referenced_model_cache() Manager->>Discovery: warm referenced providers par Provider queries Discovery->>Provider: list_model_infos() end Provider-->>Discovery: metadata or failure Discovery->>Cache: cache successful results Discovery-->>Manager: refresh result Manager-->>App: warm complete App->>Manager: start_model_list_refresh() Manager->>Discovery: refresh only missing providers Discovery->>Cache: cache remaining catalogs ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Unify provider model discovery ownership"](https://github.com/alishahryar1/free-claude-code/commit/33f68e4338fd2326a7d8fd3e7e28246b47fa4210) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45862060)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
36bb282558 |
Recover Claude sessions from provider context overflow (#1215)
## Problem NVIDIA NIM can report context exhaustion as a 400 `BadRequestError` saying its derived `max_tokens` is negative. FCC treated that as a generic invalid request, so Claude Code could not recognize the smaller upstream context window, compact the conversation, and replay the interrupted turn. LM Studio also encoded Claude's recovery phrase inside provider code instead of reporting a protocol-neutral failure. Fixes #1198. ## Changes - Add one protocol-neutral `context_window_exceeded` execution failure with a non-retryable 400 contract. - Narrowly classify only NVIDIA NIM's negative derived-`max_tokens` signature while preserving the complete redacted provider diagnostic and request ID. - Let the Anthropic serializer alone add Claude's `prompt is too long` compaction trigger; OpenAI Responses keeps a standard invalid-request envelope. - Migrate LM Studio's existing context preflight to the same neutral semantic and document the ownership boundary. - Add provider, protocol, API, trace, near-miss, and real Claude compaction/replay coverage; bump FCC to 4.11.4. | Before | After | | --- | --- | | Context overflow appeared as an ordinary provider 400 and ended the Claude turn. | Claude receives a typed 400 with its recognized compaction trigger, compacts once, and replays the interrupted turn without an FCC or SDK retry loop. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds protocol-neutral recovery from provider context-window exhaustion. The main changes are: - Classify NVIDIA NIM negative derived-`max_tokens` errors as context-window failures. - Move Claude's compaction trigger into the Anthropic serializer. - Migrate LM Studio context preflight to the neutral failure type. - Preserve the standard OpenAI Responses invalid-request envelope. - Add provider, protocol, API, trace, and recovery tests. - Bump the package version to 4.11.4. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. - No blocking issues found in the changed code. - The new failure kind is covered by both protocol mappings. - Provider classification remains narrow and non-retryable. - The protocol-specific compaction phrase stays at the Anthropic boundary. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the narrow pytest command from /home/user/repo without live provider credentials or services, and observed a clean test run with all tests passing. <a href="https://app.greptile.com/trex/runs/15073403/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/nvidia_nim/client.py | Adds narrow context-window classification for nested and top-level NVIDIA NIM error bodies. | | src/free_claude_code/providers/lmstudio/client.py | Migrates context-budget preflight failures to the canonical context-window failure. | | src/free_claude_code/core/anthropic/errors.py | Adds Anthropic mapping and injects Claude's compaction phrase at the wire boundary. | | src/free_claude_code/core/openai_responses/errors.py | Maps context exhaustion to the standard OpenAI invalid-request error. | | src/free_claude_code/providers/failure_policy.py | Adds the canonical non-retryable status-400 context-window failure factory. | | src/free_claude_code/core/failures.py | Adds the protocol-neutral context-window failure category. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart TD A[Provider detects context exhaustion] --> B[Context-window ExecutionFailure] B --> C{Protocol adapter} C -->|Anthropic Messages| D[400 invalid_request_error] D --> E[Add prompt is too long trigger] E --> F[Claude compacts and replays] C -->|OpenAI Responses| G[400 invalid_request_error] G --> H[Keep neutral provider message] ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart TD A[Provider detects context exhaustion] --> B[Context-window ExecutionFailure] B --> C{Protocol adapter} C -->|Anthropic Messages| D[400 invalid_request_error] D --> E[Add prompt is too long trigger] E --> F[Claude compacts and replays] C -->|OpenAI Responses| G[400 invalid_request_error] G --> H[Keep neutral provider message] ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Normalize provider context overflow for ..."](https://github.com/alishahryar1/free-claude-code/commit/29c89a4c1011d986284e9d163855174d1d433293) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45603631)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
fe40548a30 |
Give Google reasoning controls one owner (#1211)
## Problem Gemini requests with FCC reasoning could send both `reasoning_effort` and `extra_body.google.thinking_config`, which [Google documents as mutually exclusive](https://ai.google.dev/gemini-api/docs/openai#thinking). Google reasoning and thought-signature postprocessing had overlapping request ownership. Fixes #1206. ## Changes | Before | After | | --- | --- | | Shared Google quirks injected thought output independently of the profile encoder. | One provider-selected Google encoder owns every reasoning wire field. | | Gemini could send named effort beside a custom thinking config. | Gemini selects one channel, with exact budgets taking precedence over named effort. | | Vertex reasoning and thought-signature behavior shared one quirks module. | Vertex retains its budget mapping while thought signatures have a separate owner. | | Caller-native Google controls could collide with FCC controls. | Native controls are preserved only under provider-default reasoning; controlled collisions fail preflight. | | Regression coverage inspected isolated body fragments. | Policy matrices, SDK-merge assertions, and an ownership contract enforce the final wire shape. | | The reported Gemini 3.5 Flash path failed upstream with HTTP 400. | The same live path streams to a normal terminal stop with one reasoning channel. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR gives each Google provider one owner for reasoning request fields. The main changes are: - Adds dedicated Gemini and Vertex reasoning encoders. - Separates thought-signature replay from reasoning serialization. - Validates caller-provided Google configuration before encoding. - Adds request-policy and final wire-shape tests. - Bumps the package version to 4.11.2. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. The request pipeline keeps thought signatures and reasoning fields separate. Tests cover the supported reasoning policies and caller configuration conflicts. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The focused validation suite was executed and completed with 78 passed in 4.57s. - Before the change, adaptive/high emitted two channels: reasoning\_effort: high and thinking\_config.include\_thoughts: true. - After the change, the same request emits exactly one channel: reasoning\_effort: high, with thinking\_config: null. <a href="https://app.greptile.com/trex/runs/15056242/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/google_openai/reasoning.py | Adds exclusive Gemini and Vertex reasoning encoders and validates caller-native thinking configuration. | | src/free_claude_code/providers/google_openai/provider.py | Separates message signature replay from profile-owned reasoning encoding. | | src/free_claude_code/providers/google_openai/thought_signatures.py | Narrows the former quirks module to tool-call thought-signature replay. | | src/free_claude_code/providers/gemini/client.py | Selects the Gemini encoder and enables validated extra-body forwarding. | | src/free_claude_code/providers/vertex/client.py | Selects the Vertex encoder while retaining budget-based reasoning controls. | | tests/providers/test_gemini.py | Covers channel exclusivity, budget precedence, native configuration, conflicts, and SDK merging. | | tests/providers/test_vertex.py | Covers Vertex policy mapping, native configuration, conflicts, and final wire shape. | | tests/contracts/test_import_boundaries.py | Enforces one source owner for Google reasoning wire fields. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart TD A[Messages request] --> B[Resolve reasoning policy] A --> C[Validate and copy extra_body] B --> D{Google provider profile} C --> E[Build OpenAI request body] E --> F[Replay thought signatures] F --> D D -->|Gemini| G[Gemini reasoning encoder] D -->|Vertex| H[Vertex reasoning encoder] G --> I[One Gemini reasoning channel] H --> J[Google thinking configuration] I --> K[Final SDK request] J --> K ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart TD A[Messages request] --> B[Resolve reasoning policy] A --> C[Validate and copy extra_body] B --> D{Google provider profile} C --> E[Build OpenAI request body] E --> F[Replay thought signatures] F --> D D -->|Gemini| G[Gemini reasoning encoder] D -->|Vertex| H[Vertex reasoning encoder] G --> I[One Gemini reasoning channel] H --> J[Google thinking configuration] I --> K[Final SDK request] J --> K ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Give Google reasoning controls one owner"](https://github.com/alishahryar1/free-claude-code/commit/cdfba8273878d5075ac8d6afb702e0f76224cba4) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45569811)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
af12e7b2bb |
Coordinate provider recovery under concurrent load (#1205)
## Problem Concurrent transient failures could start independent retry, replay, continuation, and repair loops while holding provider concurrency slots. This multiplied upstream attempts and could delay or strand terminal errors under fan-out. ## Changes | Before | After | | --- | --- | | Retry paths owned separate attempt budgets. | One logical-execution session caps all upstream work at five attempts. | | Concurrent failures backed off independently. | One provider-owned recovery episode elects a single half-open probe while followers coalesce. | | Backoff occupied stream concurrency. | Concurrency is held only while an upstream operation or stream is active. | | Provider catalog calls and stream creation used separate admission paths. | Every upstream operation uses one provider-generation admission controller. | | Cancellation could leave recovery ownership or follower state unresolved. | Cancellation releases permits, transfers probe ownership, and unregisters waiting followers. | | Late in-flight failures could cross an exhausted episode boundary. | Every coalesced execution retains that generation's terminal outcome. | | Replay tests allowed loose lifecycle assertions. | Exact SSE contracts prove retries and continuations emit one unduplicated response. | | Recovery wrappers could mask final diagnostics. | Final responses and traces retain the raw provider failure and request ID. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR coordinates provider recovery and retry work under concurrent load. The main changes are: - One five-attempt budget for each logical execution. - Provider-wide recovery episodes with one elected probe. - Shared admission for streams, catalog calls, rate limits, and concurrency. - Concurrency permits held only during active upstream work. - Cancellation-safe probe ownership and preserved final diagnostics. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Reviewed the coordinated-recovery-01-before.log to understand how the exhausted generation outcome was not preserved in a late in-flight failure. - Reviewed the coordinated-recovery-02-after.log to confirm that the updated implementation preserves the exhausted generation outcome for the same focused contract set. - Validated that the provider-admission-full-current.log shows the complete requested test file passed under Python 3.14 with uv run pytest -n 0. <a href="https://app.greptile.com/trex/runs/15050270/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/admission.py | Adds shared admission, retry budgets, recovery episodes, probe election, and cancellation handling. | | src/free_claude_code/providers/openai_chat/provider.py | Moves stream creation, replay, continuation, and repair onto one admission-owned retry session. | | src/free_claude_code/providers/stream_recovery.py | Selects replay, continuation, repair, or final failure using the remaining shared attempt budget. | | src/free_claude_code/providers/failure_policy.py | Adds recovery exhaustion handling and preserves the underlying provider error for final classification. | | src/free_claude_code/providers/runtime/factory.py | Creates one admission controller per provider generation and passes it through provider factories. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant E as Execution participant A as Admission controller participant P as Provider participant F as Concurrent follower E->>A: Open attempt A->>P: Send upstream request P-->>E: Retryable failure E->>A: Open recovery episode F->>A: Request admission A-->>F: Coalesce and wait E->>A: Claim probe A->>P: Send half-open probe alt Probe succeeds P-->>E: Valid response E->>A: Close recovery episode A-->>F: Release waiter else Probe fails P-->>E: Retryable failure E->>A: Schedule next probe or finalize error end ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant E as Execution participant A as Admission controller participant P as Provider participant F as Concurrent follower E->>A: Open attempt A->>P: Send upstream request P-->>E: Retryable failure E->>A: Open recovery episode F->>A: Request admission A-->>F: Coalesce and wait E->>A: Claim probe A->>P: Send half-open probe alt Probe succeeds P-->>E: Valid response E->>A: Close recovery episode A-->>F: Release waiter else Probe fails P-->>E: Retryable failure E->>A: Schedule next probe or finalize error end ``` </a> </details> <sub>Reviews (2): Last reviewed commit: ["Harden coordinated retry lifecycle invar..."](https://github.com/alishahryar1/free-claude-code/commit/2e871c8649d148b5eb71d21f80bf870ae2d11708) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45554917)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
a0f62c598c |
Add Google Vertex AI with renewable ADC (#1193)
## Problem | Before | After | | --- | --- | | FCC supported Google AI Studio API keys but could not route coding agents through a Google Cloud Vertex AI project. | `vertex/...` routes through Google's [documented OpenAI-compatible Chat Completions endpoint](https://cloud.google.com/vertex-ai/generative-ai/docs/start/openai), using the global endpoint by default or an explicitly configured region. | | A pasted Vertex access token would expire, while Application Default Credentials were not part of provider construction. | FCC loads [Application Default Credentials](https://cloud.google.com/docs/authentication/application-default-credentials), supplies a renewable credential callback to the OpenAI transport, coalesces concurrent refreshes, and returns typed authentication or transient failures. | | Vertex does not expose its model catalog through the compatible OpenAI `/models` route. | FCC translates its generic discovery operation to Google's paginated [publisher-model list API](https://cloud.google.com/vertex-ai/docs/reference/rest/v1beta1/publishers.models/list) and converts resource names into the model IDs accepted by Chat Completions. | | Google thought signatures were owned by the AI Studio adapter even though Vertex shares the same protocol behavior. | A neutral Google OpenAI family owns shared thought-signature and request behavior; AI Studio and Vertex retain separate endpoint and authentication ownership. | ## Changes - Added the Vertex provider, `VERTEX_PROJECT_ID`, optional `VERTEX_LOCATION` and `VERTEX_PROXY`, Admin UI configuration, model-picker discovery, smoke metadata, and customer setup documentation. - Added renewable ADC access tokens with refresh coalescing, proxy-aware refresh, sanitized failure classification, and project quota headers. - Added global/regional endpoint composition plus native model-catalog pagination, strict response validation, response cleanup, and repeated-page protection. - Generalized provider readiness around declared configuration fields so project-based and multi-field providers no longer pretend every remote provider is configured by one API key. - Moved shared Google request quirks out of the Gemini adapter, preserved AI Studio behavior, and bumped the package to `4.11.0`. <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Google Vertex AI as a new provider using Application Default Credentials. The main changes are: - New `vertex` provider with project/location endpoint construction. - Renewable ADC access-token loading with refresh coalescing and proxy-aware refresh. - Native Vertex publisher-model discovery with pagination and response validation. - Shared Google OpenAI-compatible request behavior for Gemini and Vertex. - Admin UI, settings, smoke config, docs, version, lockfile, and tests for the new provider. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with low risk. No blocking correctness or security issues were identified. The new provider follows the existing provider-runtime and Admin configuration patterns. Endpoint, auth, model parsing, readiness, docs, version, lockfile, and tests are updated together. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The T-Rex test suite was executed to validate the code-execution proof-of-work, generating a full verbose pytest log and recording the run metadata, and the run completed with EXIT\_CODE: 0. <a href="https://app.greptile.com/trex/runs/14991235/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/vertex/client.py | Adds the Vertex provider with OpenAI-compatible chat routing and native paginated model discovery. | | src/free_claude_code/providers/vertex/auth.py | Implements renewable ADC token loading, proxy-aware refresh, coalescing, and sanitized auth failures. | | src/free_claude_code/providers/vertex/endpoint.py | Builds validated Vertex global/regional service, chat, and model-list endpoints. | | src/free_claude_code/providers/vertex/models.py | Parses Vertex publisher-model pages into OpenAI-compatible model IDs with malformed-response checks. | | src/free_claude_code/providers/google_openai/provider.py | Adds shared Google thought-signature caching and thinking-budget request body handling. | | src/free_claude_code/providers/google_openai/quirks.py | Renames Gemini-specific quirks to shared Google quirks and exposes model-neutral thinking config helpers. | | src/free_claude_code/providers/openai_chat/provider.py | Allows OpenAI-chat providers to pass an async API-key callback into the OpenAI SDK. | | src/free_claude_code/providers/runtime/discovery.py | Uses descriptor-defined readiness to choose providers eligible for model cache/discovery. | | src/free_claude_code/config/provider_catalog.py | Adds the Vertex descriptor and required settings metadata, and makes Cloudflare readiness require both token and account ID. | | src/free_claude_code/config/admin/status.py | Generalizes Admin provider readiness status to use each descriptor's configuration attributes. | | src/free_claude_code/config/admin/provider_manifest.py | Adds Admin UI fields for Vertex project and location alongside generated provider fields. | | tests/providers/test_vertex.py | Adds targeted tests for Vertex endpoints, ADC token refresh, reasoning mapping, and model discovery pagination. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant User as User / Admin UI participant Settings as Settings + Provider Catalog participant Runtime as Provider Runtime participant Vertex as VertexProvider participant ADC as Google ADC participant OpenAI as OpenAI-compatible Chat Endpoint participant Models as Vertex Publisher Models API User->>Settings: Set VERTEX_PROJECT_ID / VERTEX_LOCATION / VERTEX_PROXY Settings->>Runtime: Descriptor reports vertex configured by project id Runtime->>Vertex: Construct with project, location, proxy, rate limiter Vertex->>ADC: Load/refresh Application Default Credentials ADC-->>Vertex: Renewable access token Vertex->>OpenAI: Stream chat completion with bearer token + x-goog-user-project OpenAI-->>Vertex: Streaming chat chunks Vertex-->>Runtime: Normalized provider stream Runtime->>Vertex: Refresh model list Vertex->>Models: GET paginated publishers/google/models Models-->>Vertex: publisherModels + nextPageToken Vertex-->>Runtime: Prefixed model IDs for cache/model picker ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant User as User / Admin UI participant Settings as Settings + Provider Catalog participant Runtime as Provider Runtime participant Vertex as VertexProvider participant ADC as Google ADC participant OpenAI as OpenAI-compatible Chat Endpoint participant Models as Vertex Publisher Models API User->>Settings: Set VERTEX_PROJECT_ID / VERTEX_LOCATION / VERTEX_PROXY Settings->>Runtime: Descriptor reports vertex configured by project id Runtime->>Vertex: Construct with project, location, proxy, rate limiter Vertex->>ADC: Load/refresh Application Default Credentials ADC-->>Vertex: Renewable access token Vertex->>OpenAI: Stream chat completion with bearer token + x-goog-user-project OpenAI-->>Vertex: Streaming chat chunks Vertex-->>Runtime: Normalized provider stream Runtime->>Vertex: Refresh model list Vertex->>Models: GET paginated publishers/google/models Models-->>Vertex: publisherModels + nextPageToken Vertex-->>Runtime: Prefixed model IDs for cache/model picker ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["feat: add Google Vertex AI provider"](https://github.com/alishahryar1/free-claude-code/commit/97e753f0772e60377865876ca59b2fd8888d922e) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45405432)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
af658287bd |
Add Amazon Bedrock Mantle support (#1192)
## Problem FCC cannot route coding-agent requests through Amazon Bedrock even though Bedrock Mantle exposes an OpenAI-compatible streaming Chat Completions API. Users currently need a separate compatibility layer, and FCC has no catalog, Admin UI, model-discovery, or smoke-test contract for Bedrock. Fixes #863. ## Changes | Before | After | | --- | --- | | Amazon Bedrock was absent from provider routing. | `bedrock/` uses the existing OpenAI Chat provider with AWS's [Bedrock Mantle endpoint](https://docs.aws.amazon.com/bedrock/latest/userguide/inference-chat-completions-mantle.html). | | Bedrock integration would have implied a new AWS-native transport. | The ordinary profile owns streaming, tools, retries, and `/models` discovery without boto3, SigV4, Converse, or Invoke machinery. | | FCC had no Bedrock authentication or regional endpoint configuration. | `AWS_BEARER_TOKEN_BEDROCK`, `BEDROCK_BASE_URL`, and `BEDROCK_PROXY` are available through environment and Admin UI configuration, with the current `us-east-1` Mantle URL as the default. | | Heterogeneous Bedrock models had no safe provider-wide reasoning control. | FCC replays prior reasoning through portable think tags and leaves model-specific reasoning parameters upstream-owned. | | Smoke configuration duplicated credential checks for every provider. | Smoke configuration reads primary credentials and configurable endpoints from the provider catalog, retaining only Cloudflare's two-field exception. | | Bedrock behavior had no deterministic coverage or release documentation. | Provider requests, regional URL normalization, model discovery, Admin persistence, smoke selection, README usage, architecture boundaries, and version 4.10.0 cover the new capability. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Amazon Bedrock Mantle as an OpenAI-compatible provider. The main changes are: - Bedrock catalog, settings, proxy, and regional base-URL configuration. - OpenAI Chat routing with portable reasoning replay and model discovery. - Admin UI fields and configuration persistence. - Catalog-driven smoke configuration and Bedrock smoke coverage. - Provider documentation and a version bump to 4.10.0. </details> <h3>Confidence Score: 5/5</h3> The provider flow looks mergeable after handling an explicitly empty Bedrock base URL. Catalog, runtime, Admin, and smoke wiring are consistent. Regional URLs with or without `/v1` are normalized correctly. src/free_claude_code/config/settings.py: an empty `BEDROCK_BASE_URL` can still create a client with an invalid endpoint. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Before the change, bedrock was rejected as an unknown provider with exit code 1. - After the change, the fake service captured the normalized URL, Bearer trex...oken, portable Chat Completions JSON, one tool, think-tag history, and no reasoning\_effort, reasoning, or thinking fields; exit code 0. - Focused pytest validation completed with 6/6 passing tests. <a href="https://app.greptile.com/trex/runs/14988686/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/config/provider_catalog.py | Adds the Bedrock provider descriptor, regional default endpoint, credential mapping, and proxy metadata. | | src/free_claude_code/config/settings.py | Adds Bedrock settings, but an explicitly empty base URL bypasses the regional default. | | src/free_claude_code/providers/openai_chat/profiles.py | Registers Bedrock with URL normalization, think-tag replay, and no provider-wide reasoning parameter. | | src/free_claude_code/config/admin/provider_manifest.py | Generalizes provider base-URL fields and adds Bedrock-specific labels and help text. | | smoke/lib/config.py | Adds Bedrock smoke defaults and replaces provider-specific checks with catalog-driven configuration checks. | | tests/providers/test_bedrock.py | Covers Bedrock URL normalization, request fields, reasoning replay, tool calls, and model discovery. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart LR Config[Environment or Admin config] --> Settings[Settings] Catalog[Provider catalog] --> Runtime[Provider runtime] Settings --> Runtime Runtime --> Profile[Bedrock OpenAI Chat profile] Profile --> Client[AsyncOpenAI client] Client --> Mantle[Regional Bedrock Mantle endpoint] Catalog --> Admin[Admin manifest] Catalog --> Smoke[Smoke selection] ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart LR Config[Environment or Admin config] --> Settings[Settings] Catalog[Provider catalog] --> Runtime[Provider runtime] Settings --> Runtime Runtime --> Profile[Bedrock OpenAI Chat profile] Profile --> Client[AsyncOpenAI client] Client --> Mantle[Regional Bedrock Mantle endpoint] Catalog --> Admin[Admin manifest] Catalog --> Smoke[Smoke selection] ``` </a> </details> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22ali%2Fadd-bedrock-mantle%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22ali%2Fadd-bedrock-mantle%22.%0A%0AFix%20the%20following%201%20code%20review%20issue.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%201%0Asrc%2Ffree_claude_code%2Fconfig%2Fsettings.py%3A60-63%0A**Empty%20Override%20Bypasses%20Default%20URL**%0A%0AWhen%20%60BEDROCK_BASE_URL%60%20is%20present%20but%20empty%2C%20settings%20keep%20the%20empty%20string%20instead%20of%20using%20%60BEDROCK_DEFAULT_BASE%60.%20Clearing%20this%20field%20through%20environment%20or%20Admin%20configuration%20therefore%20builds%20the%20Bedrock%20client%20with%20an%20invalid%20base%20URL%2C%20and%20Bedrock%20requests%20and%20model%20discovery%20fail%20instead%20of%20using%20the%20documented%20default.%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1192&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (1): Last reviewed commit: ["Add Amazon Bedrock Mantle provider"](https://github.com/alishahryar1/free-claude-code/commit/f40b539c99c9f6a888762b8da69d5a738a4f4d70) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45400300)</sub> > Greptile also left **1 inline comment** on this PR. <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
b89e849fff |
Add Kimi Code subscription support (#1183)
## Problem FCC's existing `kimi` provider targets Kimi Open Platform credits. Kimi Code subscription keys use a separate coding-agent endpoint, so subscribers cannot currently use their plan or select its K3 and coding models. Fixes #1161. ## Changes | Before | After | | --- | --- | | `KIMI_API_KEY` was the only Kimi contract and routed to the credit-based API platform. | `KIMI_API_KEY` remains unchanged, while `KIMI_CODE_API_KEY` routes the separate [Kimi Code subscription](https://www.kimi.com/code/docs/en/) through `https://api.kimi.com/coding/v1`. | | OpenAI-chat profiles could not declare an upstream client identity. | The Kimi Code profile sends an honest `free-claude-code` user agent as required by [Kimi's integration policy](https://www.kimi.com/code/docs/en/kimi-code/community-guidelines.html), without adding a specialized provider class. | | FCC's fallback output limit and generic token field would override Kimi's default when the client omitted a limit. | Explicit client limits become `max_completion_tokens`; omitted limits remain upstream-owned, and learned cap recovery still handles model-specific rejections. | | Kimi Code reasoning and history had no provider contract. | The profile maps FCC intent to Kimi's documented `low`, `high`, `max`, and `none` efforts and replays prior thinking through `reasoning_content`, with no model-name branching. | | Admin, model discovery, and smoke coverage knew only the credit-based Kimi provider. | Catalog-derived Admin fields, `/models` discovery, README setup, smoke configuration, and deterministic provider/runtime contracts cover `kimi_code`; the release is bumped to 4.9.0. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Kimi Code subscription support as a separate provider. The main changes are: - Adds dedicated API key, proxy, endpoint, and Admin settings. - Adds Kimi-specific reasoning, token-limit, history, and user-agent behavior. - Extends model discovery and smoke configuration for `kimi_code`. - Adds provider contract tests and updates documentation. - Bumps the package and lockfile version to 4.9.0. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The contract validation compared the Kimi code HTTP flow before and after the change and confirmed the after state exits with code 0 and includes Bearer authentication, User-Agent: free-claude-code, model discovery, max\_completion\_tokens, Kimi reasoning effort max, reasoning\_content replay, and omission of token/reasoning fields when unspecified. - Focused tests for the Kimi code HTTP flow were run, and 88 tests passed with exit code 0. - Artifacts document both the initial failure scenario and the successful post-change state, along with the focused tests results. <a href="https://app.greptile.com/trex/runs/14945625/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/config/provider_catalog.py | Registers Kimi Code with its dedicated credential, endpoint, and proxy setting. | | src/free_claude_code/config/settings.py | Adds environment-backed fields for the Kimi Code subscription key and proxy. | | src/free_claude_code/providers/openai_chat/__init__.py | Passes an optional profile user agent into the shared OpenAI-compatible client. | | src/free_claude_code/providers/openai_chat/profiles.py | Defines Kimi Code reasoning, token-limit, history replay, extra-body, and user-agent behavior. | | smoke/lib/config.py | Adds Kimi Code configuration detection and its default smoke model. | | tests/providers/test_kimi_code.py | Covers the new endpoint, headers, request mapping, reasoning behavior, and model discovery. | </details> <sub>Reviews (1): Last reviewed commit: ["Add Kimi Code subscription provider"](https://github.com/alishahryar1/free-claude-code/commit/dfc66f2755e5885775a77f5c91b4b9a82eb8fd1d) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45309322)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
65a342ede4 |
Honor reasoning tiers on numeric-budget providers (#1150)
## Problem FCC preserved named reasoning effort but numeric-budget providers received no intensity unless a client supplied an exact token budget. NIM and llama.cpp therefore treated Low through Max like the same provider default. ## Changes | Before | After | | --- | --- | | Named effort had no FCC-owned numeric meaning. | FCC maps Low, Medium, High, X-High, and Max to 512, 1,024, 2,048, 4,096, and 8,192 tokens. | | NIM received only thinking booleans for named effort. | [NIM](https://docs.nvidia.com/nim/large-language-models/1.15.0/thinking-budget-control.html) receives thinking booleans plus the mapped `reasoning_budget`, retaining its existing retry without rejected budget control. | | llama.cpp forwarded only exact client budgets. | [llama.cpp](https://github.com/ggml-org/llama.cpp/blob/master/tools/server/README.md) receives the mapped `thinking_budget_tokens` value. | | Named, boolean, and provider-default adapters shared no explicit numeric contract. | Named adapters keep words, boolean adapters keep on/off, and only numeric-budget adapters consume the FCC scale. | | Documentation prohibited every named-effort budget conversion. | Documentation defines the product scale, exact-budget precedence, and model-independent ownership boundary. | | Version 4.8.0 exposed tiers that collapsed on numeric providers. | Version 4.8.1 completes the tiers; all five local CI checks pass with 2,383 tests passed and 40 skipped. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR gives named reasoning tiers numeric budgets for providers that require token counts. The main changes are: - Adds one shared tier-to-token scale with exact-budget precedence. - Sends mapped budgets to NVIDIA NIM and llama.cpp. - Removes conflicting client-supplied NIM budget fields. - Adds focused provider and policy tests. - Updates documentation and bumps the package to 4.8.1. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. The NIM fix removes both conflicting budget locations while preserving unrelated nested options. No blocking issue remains in the changed paths. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex ran a pre-change focused validation against the budget logic and observed 16 failures and 2 passes, establishing the baseline. - T-Rex ran a post-change focused validation (head run) and confirmed all 18 cases passed, including the mocked NIM retry paths. - T-Rex executed the full focused validation after the change and produced a verbose log showing 91/91 passing. <a href="https://app.greptile.com/trex/runs/14800553/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/core/reasoning.py | Adds the shared effort-to-token scale and preserves exact client budgets as the higher-priority value. | | src/free_claude_code/providers/nvidia_nim/request_options.py | Removes top-level and nested client budgets before inserting one policy-derived NIM budget. | | src/free_claude_code/providers/openai_chat/reasoning.py | Extends llama.cpp request encoding to send mapped named-effort budgets. | </details> <sub>Reviews (2): Last reviewed commit: ["fix: canonicalize NIM reasoning budgets"](https://github.com/alishahryar1/free-claude-code/commit/368f1a88d3f1404fcab2365703cca89f1077f1ea) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44998377)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
0de7608b52 |
Keep inline system reminders cache-stable across providers (#1154)
## Problem Inline Anthropic system reminders were forwarded as mid-conversation OpenAI system roles. Compatible provider chat templates could reposition those roles, changing prior prompt tokens and causing periodic full cache misses. Fixes #1152. ## Changes | Before | After | | --- | --- | | Top-level and inline system content both used downstream system roles. | Only top-level system content uses the leading downstream system role. | | Inline system reminders could trigger provider-side prompt retemplating. | Inline system reminders keep their text and position as downstream user messages. | | Provider policies could reinterpret the shared role mapping. | Shared conversion owns one provider-independent role mapping. | | Cache-prefix coverage allowed mid-conversation system roles. | Cache-prefix coverage requires append-only user-encoded reminders. | | The package version was 4.8.0. | The package version is 4.8.1. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR keeps inline system reminders stable across OpenAI-compatible providers. The main changes are: - Maps only the top-level system prompt to the downstream system role. - Encodes inline system reminders as ordered user content. - Coalesces adjacent user messages after transcript ordering. - Adds tests for text, multimodal content, cache-prefix stability, and tool-result ordering. - Updates the architecture notes and bumps the package to 4.8.1. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. Adjacent user messages are combined after assistant and tool dependencies are ordered. Multimodal content keeps its part order. No blocking issues were found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Before capture, the converted prefix roles were shown as system, user, and assistant. - After capture, the same prefix appeared, followed by a user continuation containing a second question and a system-reminder tag; all runtime assertions passed. - The preserved harness documents and reproduces the public-builder validation. <a href="https://app.greptile.com/trex/runs/14803290/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/core/anthropic/conversion.py | Maps inline system reminders to user content and coalesces adjacent user messages after transcript ordering. | | tests/providers/test_converter.py | Adds tests for inline reminders, multimodal content, cache-prefix stability, and tool-result ordering. | | ARCHITECTURE.md | Documents the shared role-mapping and adjacent-user coalescing rules. | | pyproject.toml | Bumps the package version to 4.8.1. | | uv.lock | Synchronizes the locked editable package version with 4.8.1. | </details> <sub>Reviews (2): Last reviewed commit: ["fix: coalesce adjacent provider user tur..."](https://github.com/alishahryar1/free-claude-code/commit/7af6327c79fa15c0f4922bad8864f1c6656b2814) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=45005803)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
6455c63e1d |
Make reasoning policy provider-neutral and client-aware (#1148)
## Problem FCC reduced reasoning to global and route booleans, mixing client intent, configuration, provider wire capabilities, output visibility, and history replay. That discarded named client efforts, encouraged model-name checks, and made provider behavior inconsistent. ## Changes | Before | After | | --- | --- | | Admin exposed global and route thinking toggles. | Admin exposes **Off**, **From client**, **Low**, **Medium**, **High**, **X-High**, and **Max**; Fable, Opus, Sonnet, and Haiku also expose **Inherit**. | | Request intent was repeatedly reduced to a boolean across routing and providers. | The application boundary resolves one immutable `ReasoningPolicy` with independent control, named effort, and exact positive token budget. | | Provider adapters could infer reasoning behavior from upstream model names or versions. | Provider profiles translate only documented provider-wide wire capabilities; architecture and contributor rules prohibit model-specific reasoning branches. | | Gateway reasoning controls were ad hoc. | [OpenRouter](https://openrouter.ai/docs/guides/best-practices/reasoning-tokens) and [Vercel AI Gateway](https://vercel.com/docs/ai-gateway/models-and-providers) use documented reasoning objects, including exact budgets where representable. | | Named effort forwarding was inconsistent or absent. | [Gemini](https://ai.google.dev/gemini-api/docs/openai), [Ollama](https://docs.ollama.com/api/openai-compatibility), [LM Studio](https://lmstudio.ai/changelog/lmstudio-v0.4.8), [Fireworks](https://docs.fireworks.ai/guides/querying-text-models/reasoning), [Cohere](https://docs.cohere.com/docs/compatibility-api), [Wafer](https://docs.wafer.ai/serverless/api-reference), [Groq](https://console.groq.com/docs/reasoning), [Cerebras](https://inference-docs.cerebras.ai/capabilities/reasoning), [SambaNova](https://docs.sambanova.ai/docs/api-reference/chat-completions/create-chat-based-completion), and [Mistral](https://docs.mistral.ai/studio-api/conversations/reasoning) receive their documented named vocabularies with explicit provider-owned downgrades. | | Boolean thinking controls were mixed into shared conversion. | [DeepSeek](https://api-docs.deepseek.com/guides/thinking_mode/), [Kimi](https://platform.kimi.ai/docs/guide/use-kimi-k2-thinking-model), [Z.ai](https://docs.z.ai/guides/capabilities/thinking-mode), [Cloudflare Workers AI](https://developers.cloudflare.com/changelog/post/2026-04-20-kimi-k2-6-workers-ai/), and [NVIDIA NIM](https://docs.nvidia.com/nim/large-language-models/1.15.0/thinking-budget-control.html) use provider-owned thinking-object or chat-template controls. | | Effort names and output limits could become fabricated reasoning budgets. | Exact budgets remain exact and are forwarded only through documented fields for OpenRouter, Fireworks, LM Studio, NIM, and [llama.cpp](https://github.com/ggml-org/llama.cpp/blob/master/tools/server/README.md); named efforts and output limits are never converted into token budgets. | | New-turn reasoning and prior-turn replay shared one switch. | Every profile independently declares native reasoning replay, `<think>` tag replay, provider-specific replay, or no replay; **Off** suppresses new reasoning output without corrupting required history. | | Providers without a stable generic compute control received guessed controls. | [MiniMax](https://platform.minimax.io/docs/api-reference/text-openai-api) requests split output only, while [GitHub Models](https://docs.github.com/en/rest/models/inference), [Hugging Face Inference Providers](https://huggingface.co/docs/inference-providers/en/tasks/chat-completion), Codestral, and OpenCode keep provider defaults and use only their explicit replay profile. | | OpenAI Responses effort became a lossy Anthropic thinking boolean. | Responses preserves `reasoning.effort` through `output_config`, then resolves it through the same application policy as Messages without inventing a budget. | | Legacy booleans remained the persisted contract. | FCC-owned dotenv files migrate to typed `REASONING_*` values, explicit env files receive an actionable warning, documentation describes the ownership boundary, and the package advances to 4.8.0. | | Reasoning behavior was covered by scattered boolean assertions. | New policy, routing, encoder, provider, Admin, migration, Responses, and smoke contracts pass all five local CI checks: 2,368 tests passed, 40 skipped; 92 smoke tests collect and both live config migration checks pass. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR makes reasoning policy client-aware and independent of provider model names. The main changes are: - Adds one immutable reasoning policy resolved at the application boundary. - Adds typed root and route reasoning settings with Admin UI support. - Moves wire controls and history replay behavior into provider profiles. - Migrates owned dotenv files from legacy thinking booleans. - Expands provider, routing, migration, API, and smoke coverage. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the contract-validation test suite with the specified test modules, and the tests reported 78 passed in 1.53s with exit code 0. - Reviewed the complete captured output artifact reasoning-contract-02-after.log to verify the final test outcomes and successful contract validation. <a href="https://app.greptile.com/trex/runs/14792858/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/config/env_migrations.py | Migrates legacy reasoning booleans in owned dotenv files and warns for explicit environment files. | | src/free_claude_code/application/reasoning.py | Resolves client controls and configured preferences into one provider-neutral reasoning policy. | | src/free_claude_code/application/routing.py | Carries route-level reasoning preferences into request-scoped policy resolution. | | src/free_claude_code/providers/openai_chat/reasoning.py | Provides shared provider encoders for reasoning controls and replay behavior. | </details> <sub>Reviews (2): Last reviewed commit: ["chore: release reasoning controls as 4.8..."](https://github.com/alishahryar1/free-claude-code/commit/9d4be767f7dbdca5709474012f43dcdc6f4347e3) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44984039)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
cba7ed23c5 |
Fix Cerebras reasoning serialization (#1138)
## Problem Cerebras rejects multi-turn Claude conversations because FCC serializes prior assistant thinking as the unsupported `reasoning_content` field. Cerebras also streams reasoning through `delta.reasoning`, so FCC can miss reasoning output. Fixes #1136. ## Changes | Before | After | | --- | --- | | Cerebras replayed assistant thinking through `reasoning_content`. | Cerebras replays assistant thinking as `<think>`-tagged assistant content. | | Cerebras parsed streamed reasoning from `delta.reasoning_content`. | Cerebras parses streamed reasoning from `delta.reasoning`. | | The disabled-thinking test checked an impossible message role. | Regression tests inspect serialized fields, preserve tool history, and cover disabled thinking. | | The package version was `4.7.1`. | The package version is `4.7.2`. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR fixes reasoning serialization for Cerebras conversations. The main changes are: - Replays prior assistant reasoning as `<think>`-tagged content. - Reads streamed reasoning from `delta.reasoning`. - Adds tests for tool history and disabled thinking. - Bumps the package version to `4.7.2`. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. - No blocking issues found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The pre-change run showed HEAD^ failed the tagged-reasoning replay and delta.reasoning streaming tests, with 2 failures and 9 successes. - The post-change head checkout completed successfully, with 11 tests passing in 0.73s and exit code 0. <a href="https://app.greptile.com/trex/runs/14672360/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/openai_chat/profiles.py | Updates Cerebras reasoning replay and streaming field configuration. | | tests/providers/test_cerebras.py | Covers tagged reasoning replay, tool history, disabled thinking, and streamed reasoning. | | pyproject.toml | Bumps the package patch version to 4.7.2. | | uv.lock | Synchronizes the locked editable package version. | </details> <sub>Reviews (1): Last reviewed commit: ["Fix Cerebras reasoning serialization"](https://github.com/alishahryar1/free-claude-code/commit/bff27ef4de4c968408e6a049bc6cca2068c7c836) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44744399)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
8d7f560589 |
Preserve DeepSeek cache prefixes across tool turns (#1126)
## Problem DeepSeek cache accounting was restored, but a later tool turn could still make FCC rewrite older assistant history. That prevented FCC from guaranteeing the identical serialized prefix required for cache reuse and left the remaining behavior in #904 unresolved. ## Changes | Before | After | | --- | --- | | Non-tool reasoning was replayed until a later tool fallback removed it. | Non-tool reasoning is omitted consistently from its first history serialization. | | Current-generation thinking controlled all historical reasoning replay. | Required tool-call reasoning is replayed independently of current-generation thinking. | | One replayable tool turn made the entire tool history appear safe. | Every tool-call turn must have replayable reasoning before thinking remains enabled. | | Prefix behavior was not verified after SDK request serialization. | DeepSeek tests capture final SDK JSON and require an exact append-only message prefix. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR preserves DeepSeek cache prefixes while retaining required tool-call reasoning. The main changes are: - Omits non-tool reasoning from its first history serialization. - Replays tool-call reasoning independently of current thinking mode. - Disables thinking unless every historical tool call has replayable reasoning. - Tests the final SDK-serialized message prefix. - Bumps the package patch version and updates the lockfile. </details> <h3>Confidence Score: 4/5</h3> The reasoning-only assistant path needs a fix before merging. The main prefix-preservation flow is covered at the SDK wire boundary. Filtering a `redacted_thinking`-only turn emits empty assistant content instead of the converter's non-empty sentinel. Other request-policy callers retain their previous behavior. src/free_claude_code/providers/deepseek/compat.py <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex executed the requested verification of the code base. - T-Rex ran the focused test wire\_messages\_keep\_prefix\_across\_tool\_thinking\_fallback with pytest; the exact command and harness are captured in the log, and the test completed with exit code 0. <a href="https://app.greptile.com/trex/runs/14533439/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/deepseek/compat.py | Separates historical tool reasoning from current thinking, but can serialize an empty assistant message after filtering. | | src/free_claude_code/providers/openai_chat/request_policy.py | Adds an optional reasoning-history switch while preserving existing defaults for other callers. | | tests/providers/test_deepseek.py | Adds SDK wire-format tests for stable prefixes and independent tool-reasoning replay. | | pyproject.toml | Bumps the package patch version to 4.6.4. | | uv.lock | Synchronizes the editable package version with the project metadata. | | ARCHITECTURE.md | Documents DeepSeek's per-turn reasoning replay and append-only prefix behavior. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart LR A[Anthropic history] --> B[Classify tool-call reasoning] B --> C[Sanitize assistant history] C --> D[Convert to OpenAI messages] D --> E[Serialize through SDK] E --> F[DeepSeek request] B -->|Every tool call replayable| G[Keep current thinking] B -->|Any tool call not replayable| H[Disable current thinking] C -->|Tool-call turn| I[Retain reasoning] C -->|Non-tool turn| J[Omit reasoning] ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart LR A[Anthropic history] --> B[Classify tool-call reasoning] B --> C[Sanitize assistant history] C --> D[Convert to OpenAI messages] D --> E[Serialize through SDK] E --> F[DeepSeek request] B -->|Every tool call replayable| G[Keep current thinking] B -->|Any tool call not replayable| H[Disable current thinking] C -->|Tool-call turn| I[Retain reasoning] C -->|Non-tool turn| J[Omit reasoning] ``` </a> </details> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22ali%2Fdeepseek-cache-stable-history%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22ali%2Fdeepseek-cache-stable-history%22.%0A%0AFix%20the%20following%201%20code%20review%20issue.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%201%0Asrc%2Ffree_claude_code%2Fproviders%2Fdeepseek%2Fcompat.py%3A146%0A**Empty%20Assistant%20Content%20Reaches%20Wire**%0A%0AWhen%20an%20assistant%20turn%20contains%20only%20%60redacted_thinking%60%2C%20filtering%20removes%20every%20block%20and%20replaces%20the%20list%20with%20%60%22%22%60.%20Revalidation%20then%20takes%20the%20string-content%20path%20and%20bypasses%20the%20converter's%20existing%20%60%22%20%22%60%20fallback%2C%20so%20DeepSeek%20can%20reject%20the%20continued%20conversation%20as%20an%20empty%20assistant%20message.%0A%0A%60%60%60suggestion%0A%20%20%20%20%20%20%20%20new_msg%5B%22content%22%5D%20%3D%20filtered%20or%20%22%20%22%0A%60%60%60%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1126&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (1): Last reviewed commit: ["Preserve DeepSeek prompt cache prefixes"](https://github.com/alishahryar1/free-claude-code/commit/b07ecf28a193e9cff1624e4ef75f38344a146d49) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44440120)</sub> > Greptile also left **1 inline comment** on this PR. <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
f0b31065ee |
Preserve mid-conversation system messages through provider conversion (#1125)
## Problem FCC hoisted inline Anthropic `system` messages into the top-level system prompt during request validation. Mid-conversation system messages are position-sensitive, so this applied later instructions retroactively, changed the existing prompt/cache prefix, and prevented provider conversion from seeing the original transcript. ## Changes - Preserve inline `system` messages, content, metadata, and ordering in Messages and token-count requests while keeping the top-level system prompt distinct. - Convert text-only inline system messages to OpenAI Chat `system` messages at the same transcript position; reject unrepresentable inline blocks before streaming instead of silently dropping them. - Remove the lossy normalization path and its unused role enum, and document protocol-model versus target-conversion ownership in `ARCHITECTURE.md`. - Cover API routing, model serialization, cache-prefix stability, text blocks, tool-result ordering, invalid content, and token counting; bump the package to `4.6.2`. - Verify all five local CI checks (2,287 tests) and the ordered transcript against NVIDIA NIM, OpenRouter, Gemini, DeepSeek, Mistral, and Hugging Face. <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR preserves inline Anthropic system messages through provider conversion. The main changes are: - Keeps top-level and inline system content separate and ordered. - Converts text-only inline system messages without moving them. - Rejects system blocks that OpenAI Chat cannot represent safely. - Updates request detection to ignore system context when counting user turns. - Adds serialization, routing, token-counting, and conversion coverage. - Updates the package version and architecture documentation. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. The leading-system detection path ignores system entries when counting user turns. Inline system content remains ordered for provider conversion. Unsupported system blocks fail explicitly instead of being dropped. No blocking issues were found in the changed code. No files require attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Validated that the transcript roles now follow the order user, assistant, system, user and that the top-level prompt remains separate. - Verified that inline system content is no longer counted in message tokens and that cache\_control metadata survives parsing. - Confirmed that the converted OpenAI transcript preserves position and cache prefix. - Observed that a system message following a tool result is converted as assistant, tool, system. - Ran the focused pytest and confirmed 209 passed in 2.82s with exit code 0. <a href="https://app.greptile.com/trex/runs/14526513/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/core/anthropic/models.py | Preserves system-role messages in the original transcript instead of hoisting them into the top-level prompt. | | src/free_claude_code/core/anthropic/conversion.py | Converts ordered text-only system messages and rejects unsupported system content before streaming. | | src/free_claude_code/api/detection.py | Builds a read-only semantic view of system context and conversational user turns for local request detection. | </details> <sub>Reviews (2): Last reviewed commit: ["Restore optimizations with inline system..."](https://github.com/alishahryar1/free-claude-code/commit/6605ede7f604053381552106489dbd16bcd37987) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44423885)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
a092455b54 |
Add searchable model selection to the Admin UI (#1121)
## Problem Admin model routing fields required users to construct provider-prefixed model slugs. Optional tier overrides represented inheritance as an unexplained blank value. ## Changes | Before | After | | --- | --- | | Model inputs only gained suggestions after an individual provider refresh. | Model inputs load configured and discovered canonical slugs from one Admin catalog. | | Model routing looked like unrestricted text entry. | Model routing uses the browser's searchable model dropdown while retaining manual entry. | | Tier overrides displayed an empty value for fallback routing. | Tier overrides display **None** and persist it as an unset override. | | Model refresh returned provider-shaped cache internals. | Model refresh returns the same canonical catalog consumed by the Admin UI. | | Users inferred the provider/model slug format from examples. | The Admin UI and README define and present complete provider/model slugs. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds searchable model selection to the Admin UI. The main changes are: - Adds a canonical catalog of configured and discovered model slugs. - Adds searchable model inputs while preserving manual entry. - Represents unset tier overrides as **None**. - Reports provider-specific model refresh failures. - Reconciles cached models when provider settings change. - Updates documentation, package metadata, and tests. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. Catalog failures no longer stop the rest of the Admin UI from loading. Partial provider refreshes now produce a visible warning. Removed credential-backed providers are pruned from the shared cache, and later stale writes are rejected. No blocking issues remain in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex ran the requested verification for the pull request checks. - The verification completed, but local artifact references were not uploaded. <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/api/admin_static/admin.js | Adds searchable model fields, optional catalog hydration, refresh warnings, and None-to-unset conversion. | | src/free_claude_code/api/admin_routes.py | Adds canonical model catalog endpoints and provider refresh failure metadata. | | src/free_claude_code/providers/runtime/discovery.py | Tracks provider refresh outcomes and separates cache eligibility from discovery eligibility. | | src/free_claude_code/providers/runtime/model_cache.py | Scopes cached model metadata to currently available providers and removes stale remote entries. | | src/free_claude_code/runtime/provider_manager.py | Reconciles cache scope during runtime replacement and returns explicit refresh results. | </details> <sub>Reviews (2): Last reviewed commit: ["Fix model catalog refresh lifecycle"](https://github.com/alishahryar1/free-claude-code/commit/d16e170055f5389e538e18dece269a5f7a8c599d) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44373795)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
b1877a4b21 |
Add Ollama Cloud as a first-class provider (#1106)
## Problem FCC supports Ollama only through a local daemon, so users cannot authenticate directly to Ollama Cloud or discover its hosted models from the Admin UI. Ollama's OpenAI-compatible API also uses the standard `reasoning` field instead of `reasoning_content`, which would otherwise drop thinking output and tool-history reasoning. ## Changes | Before | After | | --- | --- | | `ollama/...` requires a local Ollama server. | `ollama_cloud/...` connects directly to `https://ollama.com/v1` with `OLLAMA_API_KEY`, while local Ollama remains unchanged. | | OpenAI-chat reasoning was hard-coded to `reasoning_content`. | Provider profiles declare their reasoning field, so Ollama streams and replays `reasoning` without a specialized transport. | | Ollama Cloud was absent from configuration, model discovery, docs, and smoke coverage. | The catalog, Admin UI, proxy setting, model picker, README, smoke matrix, and `4.5.0` release metadata expose the provider consistently. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Ollama Cloud as a separate OpenAI-compatible provider. The main changes are: - New `ollama_cloud` catalog entry, settings, admin fields, proxy setting, and smoke configuration. - Provider profiles now choose the streamed and replayed reasoning field per provider. - Ollama Cloud uses `reasoning` and `reasoning_effort`, while local Ollama stays on its separate local configuration. - Streaming recovery now respects the resolved thinking setting when collecting reasoning. - Tests, docs, environment examples, and release metadata were updated for the new provider. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. No files need attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the provider/runtime/converter/streaming tests with full verbose pytest output and confirmed EXIT\_CODE: 0. - Ran the config/catalog/contracts/admin tests with full verbose pytest output and confirmed EXIT\_CODE: 0. - Generated and ran an introspection script to verify catalog, profile, settings, and admin manifest values without external API calls, and recorded the local execution output. <a href="https://app.greptile.com/trex/runs/14367965/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/openai_chat/profiles.py | Adds provider-level reasoning field selection and the Ollama Cloud profile. | | src/free_claude_code/providers/openai_chat/provider.py | Uses the profile reasoning field for streaming and passes the thinking setting into recovery. | | src/free_claude_code/providers/openai_chat/request_policy.py | Selects reasoning replay mode from the provider policy only when thinking is enabled. | | src/free_claude_code/core/anthropic/conversion.py | Supports replaying assistant reasoning through either `reasoning_content` or `reasoning`. | | src/free_claude_code/config/provider_catalog.py | Registers Ollama Cloud as a remote provider distinct from local Ollama. | | src/free_claude_code/config/settings.py | Adds the Ollama Cloud API key and proxy settings. | </details> <sub>Reviews (3): Last reviewed commit: ["Keep local Ollama wire behavior unchange..."](https://github.com/alishahryar1/free-claude-code/commit/d6f97cdb0074a4cbda140d484fb6dd679a1406df) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=44089262)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
4f56a6aa17 |
Add Fable as a first-class Claude routing tier (#1099)
## Problem Claude Code now sends `claude-fable-5` for the Fable alias, but FCC treated it as an unrecognized model and collapsed it into the global fallback route. Users could not map Fable traffic or reasoning behavior independently. Fixes #1097. ## Changes | Before | After | | --- | --- | | Fable requests inherited `MODEL` and `ENABLE_MODEL_THINKING`. | Fable requests use `MODEL_FABLE` and `ENABLE_FABLE_THINKING` when configured, otherwise inherit the existing defaults. | | `/v1/models` omitted Claude Fable 5. | `/v1/models` advertises the canonical `claude-fable-5` identifier. | | Admin, documentation, validation, and smoke contracts described three Claude tiers. | Admin, documentation, validation, and smoke contracts describe Fable alongside Opus, Sonnet, and Haiku. | | The package version was `4.3.1`. | The package version is `4.4.0`. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Fable as a Claude routing tier. The main changes are: - `MODEL_FABLE` and `ENABLE_FABLE_THINKING` settings. - Fable routing and thinking resolution in `ModelRouter`. - `claude-fable-5` in the model catalog. - Admin, docs, smoke, and test coverage updates. - Package version bump to `4.4.0`. </details> <h3>Confidence Score: 4/5</h3> The changed routing path needs a fix for direct provider model ids containing `fable`. Fable settings, validation, admin fields, and model listing are consistent with the existing tier patterns. Blank Fable settings inherit the existing defaults. Direct provider model ids can receive the Fable thinking override when their model name contains `fable`. src/free_claude_code/application/routing.py <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Reproduced the Fable thinking overmatch by running a focused Python repro that disables global thinking and enables Fable thinking, then resolves sambanova/my-fable-ensemble-v2. - The repro confirmed direct routing preserved provider\_id=sambanova, provider\_model=my-fable-ensemble-v2, and provider\_model\_ref=sambanova/my-fable-ensemble-v2, with resolved\_thinking\_enabled and thinking\_enabled both true. - Ran the fable-tier validation pytest, which finished with exit code 0 and 177 tests passed. - Ran the runtime probe to exercise the API model list, Settings env parsing, and ModelRouter paths, and observed a 200 OK on GET /v1/models, with the catalog item id claude-fable-5 and the expected Fable vs global-default routing behavior. - Generated the probe source file used to exercise the API and Settings paths, enabling repeatable validation without real provider credentials. <a href="https://app.greptile.com/trex/runs/14275379/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/application/routing.py | Adds Fable model and thinking branches; the thinking branch can also match unrelated direct provider model ids containing `fable`. | | src/free_claude_code/config/settings.py | Adds optional Fable model and thinking settings with blank-env inheritance and provider/model validation. | | src/free_claude_code/config/model_refs.py | Includes Fable in configured chat model reference collection and dedupe. | | src/free_claude_code/config/admin/manifest.py | Adds Fable model and thinking controls to the admin manifest. | | src/free_claude_code/api/model_catalog.py | Adds `claude-fable-5` to the advertised Claude model aliases. | | pyproject.toml | Bumps the package version to `4.4.0`. | | uv.lock | Updates the editable package version to match `pyproject.toml`. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart TD A[Incoming model name] --> B{Direct provider or gateway id?} B -- yes --> C[Use provider/model directly] C --> D[Resolve thinking from provider model string] B -- no --> E{Claude tier match} E -- Fable --> F[MODEL_FABLE or MODEL] E -- Opus/Sonnet/Haiku --> G[Tier override or MODEL] E -- None --> H[MODEL] D --> I[Provider request] F --> I G --> I H --> I ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart TD A[Incoming model name] --> B{Direct provider or gateway id?} B -- yes --> C[Use provider/model directly] C --> D[Resolve thinking from provider model string] B -- no --> E{Claude tier match} E -- Fable --> F[MODEL_FABLE or MODEL] E -- Opus/Sonnet/Haiku --> G[Tier override or MODEL] E -- None --> H[MODEL] D --> I[Provider request] F --> I G --> I H --> I ``` </a> </details> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22ali%2Fadd-fable-routing-tier%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22ali%2Fadd-fable-routing-tier%22.%0A%0AFix%20the%20following%201%20code%20review%20issue.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%201%0Asrc%2Ffree_claude_code%2Fapplication%2Frouting.py%3A134-135%0A**Fable%20Thinking%20Overmatches%20Models**%0A%0AWhen%20a%20direct%20provider%20request%20uses%20a%20model%20id%20like%20%60sambanova%2Fmy-fable-ensemble-v2%60%2C%20the%20direct%20route%20bypasses%20tier%20remapping%20but%20still%20calls%20%60_resolve_thinking%28%29%60%20with%20the%20provider%20model%20string.%20With%20%60ENABLE_FABLE_THINKING%60%20set%2C%20this%20substring%20check%20applies%20Fable%20thinking%20behavior%20to%20an%20unrelated%20provider%20model%2C%20changing%20the%20outgoing%20request%20shape%20just%20because%20the%20model%20id%20contains%20%60fable%60.%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1099&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (1): Last reviewed commit: ["Add Fable as a first-class routing tier"](https://github.com/alishahryar1/free-claude-code/commit/0705840c65511fd85c74fd2c62ba8ea97afe7c12) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43892323)</sub> > Greptile also left **1 inline comment** on this PR. <!-- /greptile_comment --> |
||
|
Fethi Amari
|
1b4c7e6779 |
feat: add vision support for OpenAI chat conversion (#1061)
## Problem FCC rejects Anthropic user image blocks before vision-capable OpenAI-compatible providers can receive them. This blocks pasted images in Claude Code and IDE clients. Fixes #512. ## Changes | Before | After | | --- | --- | | User image blocks fail during Anthropic-to-OpenAI conversion. | Base64 and URL image sources become OpenAI `image_url` parts. | | Mixed text and image input cannot reach an upstream model. | Mixed content preserves exact block order and text-only behavior. | | Invalid image sources require ambiguous downstream handling. | Missing or unsupported source fields fail explicitly before execution. | | Vision support could require flags or proxy-side downloads. | Conversion stays provider-neutral and performs no network I/O. | | The package remains on `4.1.0`. | The package advances to `4.2.0` with a regenerated lockfile. | | Converter coverage rejects every user image. | Tests cover source mapping, ordering, tool boundaries, validation, and request construction. | <!-- greptile_comment --> <h3>Greptile Summary</h3> This PR adds OpenAI-compatible vision conversion for Anthropic image blocks. The main changes are: - Converts base64 image sources into data image URLs. - Converts URL image sources into OpenAI `image_url` parts. - Preserves mixed text and image ordering in user messages. - Rejects unsupported or incomplete image sources. - Updates converter tests for image and tool-result cases. - Bumps package metadata to `4.2.0` in both project and lock files. <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. No files need attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Reviewed the initial focused run and confirmed all tests passed, with the wrapper exit code reported as nonzero and retained for traceability. - Validated the focused rerun completed cleanly with 10 tests passed and exit code 0. - Validated the full converter run completed with 73 tests passed and exit code 0. <a href="https://app.greptile.com/trex/runs/14209442/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <h3>Important Files Changed</h3> | Filename | Overview | |----------|----------| | src/free_claude_code/core/anthropic/conversion.py | Adds ordered image conversion for OpenAI-compatible chat content. | | tests/providers/test_converter.py | Adds tests for image conversion, ordering, validation, and request-body output. | | pyproject.toml | Updates the package version to `4.2.0`. | | uv.lock | Updates the locked local package version to `4.2.0`. | | ARCHITECTURE.md | Documents core image conversion behavior for OpenAI-compatible providers. | | README.md | Mentions image input support across compatible models. | <sub>Reviews (8): Last reviewed commit: ["feat: add vision support for OpenAI chat..."](https://github.com/alishahryar1/free-claude-code/commit/4a6b6de14d19ce22e3be5981f84bb6f236a11181) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43544835)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
f2f5b08714 |
Make provider retry tests deterministic (#1070)
## Problem Provider retry integration tests mocked `asyncio.sleep` while reactive deadlines still used `time.monotonic`. Serial runs busy-waited through real backoffs and could exceed validation timeouts despite passing. ## Changes | Before | After | | --- | --- | | Tests globally mocked sleep without advancing the limiter clock. | A deterministic test limiter runs real retry classification and attempt accounting with zero delay. | | Provider integration tests also exercised reactive timing already covered by limiter tests. | Provider integration tests focus on retry and terminal-error contracts; limiter timing remains in its dedicated suite. | | Serial retry validation could take minutes. | The complete retry file finishes in about two seconds including startup. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR makes provider retry tests run without wall-clock backoff. The main changes are: - Adds an immediate retry limiter for provider tests. - Forces retry backoff parameters to zero in that test limiter. - Skips reactive block timing in provider retry tests. - Removes broad `asyncio.sleep` mocks from the OpenAI-compatible retry tests. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex executed the provider retry test and captured a log showing the exact command, working directory, timestamp, Pytest output, elapsed time, and exit code. <a href="https://app.greptile.com/trex/runs/14126948/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | tests/providers/support.py | Adds a test-only limiter that keeps retry classification and attempt accounting while removing elapsed-time behavior. | | tests/providers/test_openai_compat_5xx_retry.py | Removes sleep mocks from retry tests that now rely on the immediate retry limiter. | </details> <sub>Reviews (1): Last reviewed commit: ["Make provider retry tests deterministic"](https://github.com/alishahryar1/free-claude-code/commit/e5f8c05ec5900586df8d4e1e7689288f38eeab15) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43584603)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
f3ea35a777 |
Consolidate OpenAI-compatible provider adapters (#1069)
## Problem Sixteen OpenAI-compatible providers were represented by configuration-only subclasses and one factory function each. The sole provider family also lived under a multi-transport namespace that no longer described the codebase, while provider defaults, IDs, and instance caching passed through redundant forwarding layers. ## Changes | Before | After | | --- | --- | | Sixteen provider IDs used configuration-only subclasses. | Immutable OpenAI-chat profiles configure one concrete provider while preserving each provider's request policy. | | Every provider required a dedicated factory function. | Generic profile construction is the default; only eight adapters with real state or algorithms retain factories. | | The sole provider family lived under `providers/transports/`. | `providers/openai_chat/` directly owns shared request, stream, recovery, tool, and usage behavior. | | Provider defaults and IDs passed through forwarding modules, and constructors repeated fallback resolution. | The neutral catalog resolves complete immutable provider configuration once. | | `ProviderRuntime` wrapped a pass-through provider cache. | `ProviderRuntime` directly owns lazy provider instances and cleanup. | | Production carried the duplicated adapter structure. | The final shape removes a net 873 production lines with no compatibility shim or customer-facing provider change. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR consolidates OpenAI-compatible providers behind shared profiles and runtime construction. The main changes are: - OpenAI-chat behavior moved into `providers/openai_chat`. - Configuration-only provider subclasses replaced by immutable profiles. - Provider runtime now owns lazy instance caching and cleanup directly. - Provider defaults and IDs now resolve through the neutral catalog. - Tests and smoke helpers updated for the new provider shape. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. No files need attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex produced a proof for a posted P1 finding and linked it to the corresponding review comment for details. - T-Rex saved contract-validation logs for provider consolidation and preserved four log files that show the test hanging boundary and the timeout exit code 124. <a href="https://app.greptile.com/trex/runs/14125428/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/runtime/config.py | Builds resolved provider configuration from catalog descriptors, including static credentials for local providers. | | src/free_claude_code/providers/openai_chat/provider.py | Creates shared OpenAI-chat clients from resolved configuration and immutable provider profiles. | | src/free_claude_code/providers/openai_chat/profiles.py | Defines declarative profiles for formerly configuration-only OpenAI-compatible adapters. | </details> <!-- greptile_failed_comments --> <h3>Comments Outside Diff (2)</h3> 1. General comment <a href="#"><img alt="P1" src="https://greptile-static-assets.s3.amazonaws.com/badges/p1.svg?v=9" align="top"></a> **NIM OpenAI-compatible 504 exhausted streaming retry path hangs instead of completing with the expected user-facing error** - **Bug** - The broader provider consolidation suite failed all parametrized `test_nim_stream_openai_5xx_exhausted_emits_user_message` cases under xdist with worker crashes. A serial isolation run narrowed this to the same OpenAI-compatible NIM exhausted 5xx streaming path: 500, 502, and 503 completed, but the 504 case did not finish before the explicit 120s timeout, producing exit code 124. This contradicts the expected streaming retry contract that exhausted transient provider failures terminate and emit the configured user-facing error. - **Cause** - The OpenAI-compatible streaming retry/exhaustion path for NIM 504 responses appears to wait indefinitely or otherwise fail to terminate after retries are exhausted. The exact code-level loop/await point was not isolated within the validation budget, but the failure is anchored to the provider streaming retry contract exercised by `tests/providers/test_openai_compat_5xx_retry.py::test_nim_stream_openai_5xx_exhausted_emits_user_message[504-temporarily unavailable]`. - **Fix** - Inspect the OpenAI-compatible/NIM streaming retry exhaustion handling for 504 responses and ensure retry limits are enforced, the async stream is closed/cancelled on exhaustion, and the provider raises/emits the same terminal user-facing error contract as the 500/502/503 cases. Add or keep a serial regression test for the 504 exhausted stream path to prevent hangs. <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> 2. General comment <a href="#"><img alt="P1" src="https://greptile-static-assets.s3.amazonaws.com/badges/p1.svg?v=9" align="top"></a> **OpenAI-compatible 5xx retry suite hangs on exhausted 502 retry path** - **Bug** - The recommended non-xdist retry validation does not complete. A verbose rerun with a 60 second timeout shows the suite passes tests through the exhausted 500 case, then times out while running `test_nim_stream_openai_5xx_exhausted_emits_user_message[502-temporarily unavailable]`. This indicates the consolidated OpenAI-compatible provider retry/error path can hang for at least the 502 exhausted-error scenario, preventing reliable validation and potentially blocking callers from receiving the expected `ExecutionFailure`. - **Cause** - The OpenAI-compatible provider's exhausted 502 retry/error handling path appears not to terminate promptly under the mocked repeated `openai.InternalServerError` scenario. The exact code location was not changed during validation, but the failure is isolated to the shared OpenAI-chat 5xx retry behavior exercised by `NvidiaNimProvider.stream_response`. - **Fix** - Debug the shared OpenAI-chat retry loop/error classification for 502 responses. Ensure retries are bounded, patched `asyncio.sleep` is awaited without real backoff during tests, and exhausted 502/503/504 errors consistently raise `ExecutionFailure` with the expected temporary-unavailable message instead of continuing work indefinitely. <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> <!-- /greptile_failed_comments --> <sub>Reviews (2): Last reviewed commit: ["Prove local provider credential resoluti..."](https://github.com/alishahryar1/free-claude-code/commit/ba226d97efdae1889bea0372c7bacf268122d65c) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43580983)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
3a7e0ccf7a |
Remove the native Anthropic provider transport (#1067)
## Problem Ollama and llama.cpp still used a parallel native Anthropic transport after the other providers moved to OpenAI Chat. That kept duplicate request, SSE, recovery, model-list, and server-tool policy machinery alive. ## Changes | Before | After | | --- | --- | | Ollama and llama.cpp streamed through provider-specific Anthropic `/messages` adapters. | Ollama and llama.cpp use the shared OpenAI Chat transport. | | Native request serialization, SSE normalization, error mapping, and recovery remained beside the OpenAI path. | Native-only machinery is removed and all providers share one transport lifecycle. | | Routed models carried a capability object solely to permit native server-tool passthrough. | Routing carries only route decisions; FCC handles forced server tools locally and rejects lossy passthrough. | | Ollama discovery used a separate `/api/tags` parser and rejected `/v1` configuration. | Ollama discovery uses `/v1/models` and accepts either root or `/v1` base URLs. | | Obsolete native tests and a compatibility facade kept deleted internals represented. | Tests cover the shared transport and real Ollama product path without compatibility shims. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR removes the native Anthropic transport path for local providers. The main changes are: - Ollama and llama.cpp now use the shared OpenAI Chat transport. - Local provider base URLs are normalized to the OpenAI-compatible `/v1` API root. - Ollama discovery now uses the OpenAI-compatible model-listing path. - Native Anthropic transport code and server-tool passthrough capability metadata were removed. - Tests and smoke coverage were updated for the shared transport path. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the targeted local provider pytest slice and observed exit code 0. - Executed the generated runtime harness to emulate the provider HTTP interactions and capture a request trace. - Validated the request trace showed two GET /v1/models calls authorized as Bearer ollama for root and /v1 base URL configurations, and a POST /v1/chat/completions authorized as Bearer llamacpp with streaming OpenAI chat JSON payload. - Confirmed the exact generated harness script used for the runtime proof is the harness file referenced in the artifacts. <a href="https://app.greptile.com/trex/runs/14121848/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/transports/openai_chat/base_url.py | Adds a helper that normalizes local OpenAI-compatible server roots to `/v1`. | | src/free_claude_code/providers/llamacpp/client.py | Moves llama.cpp to the shared OpenAI Chat transport with local base URL normalization. | | src/free_claude_code/providers/ollama/client.py | Moves Ollama to the shared OpenAI Chat transport with local base URL normalization. | | src/free_claude_code/api/handlers/messages.py | Applies server-tool rejection through the shared request policy instead of provider passthrough metadata. | | src/free_claude_code/application/routing.py | Removes provider capability metadata from routed model results. | </details> <!-- greptile_failed_comments --> <h3>Comments Outside Diff (1)</h3> 1. `src/free_claude_code/api/handlers/messages.py`, line 261-267 ([link](https://github.com/alishahryar1/free-claude-code/blob/3c7ff176da46560c4d27b3846dca1ab1c7db561c/src/free_claude_code/api/handlers/messages.py#L261-L267)) <a href="#"><img alt="P2" src="https://greptile-static-assets.s3.amazonaws.com/badges/p2.svg?v=9" align="top"></a> **Native Server Tools Always Reject** With the passthrough capability check removed, Ollama and llama.cpp requests that previously used their native Anthropic transport for `web_search` or `web_fetch` are rejected before provider execution. The default `ENABLE_WEB_SERVER_TOOLS=false` now makes forced server-tool requests return an invalid-request error instead of reaching the local provider path that used to support them. <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22ali%2Fremove-native-anthropic-transport%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22ali%2Fremove-native-anthropic-transport%22.%0A%0AThis%20is%20a%20comment%20left%20during%20a%20code%20review.%0APath%3A%20src%2Ffree_claude_code%2Fapi%2Fhandlers%2Fmessages.py%0ALine%3A%20261-267%0A%0AComment%3A%0A**Native%20Server%20Tools%20Always%20Reject**%0A%0AWith%20the%20passthrough%20capability%20check%20removed%2C%20Ollama%20and%20llama.cpp%20requests%20that%20previously%20used%20their%20native%20Anthropic%20transport%20for%20%60web_search%60%20or%20%60web_fetch%60%20are%20rejected%20before%20provider%20execution.%20The%20default%20%60ENABLE_WEB_SERVER_TOOLS%3Dfalse%60%20now%20makes%20forced%20server-tool%20requests%20return%20an%20invalid-request%20error%20instead%20of%20reaching%20the%20local%20provider%20path%20that%20used%20to%20support%20them.%0A%0AHow%20can%20I%20resolve%20this%3F%20If%20you%20propose%20a%20fix%2C%20please%20make%20it%20concise.&repo=alishahryar1%2Ffree-claude-code&pr=1067&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixInCodex.svg?v=6"><img alt="Fix in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixInCodex.svg?v=6"></picture></a> <!-- /greptile_failed_comments --> <sub>Reviews (2): Last reviewed commit: ["Normalize local OpenAI v1 base URLs"](https://github.com/alishahryar1/free-claude-code/commit/2355eac247a6e411f89a781f46e784636ced98d6) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43572841)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
df038c2470 |
Retry degraded NVIDIA NIM functions as overloads (#1064)
## Problem NVIDIA NVCF can return HTTP 400 when a deployed function is DEGRADED, even though the failure is transient. FCC treated that response as a deterministic invalid request and never used its provider-owned retry budget. Fixes #1057. ## Changes | Before | After | | --- | --- | | Every NVIDIA HTTP 400 was non-retryable. | The NVIDIA adapter recognizes only the structured NVCF `Function id …: DEGRADED function cannot be invoked` response. | | A degraded function failed after one attempt. | The shared limiter applies its existing five-attempt exponential backoff without mutating the request or adding another retry loop. | | Exhaustion returned `INVALID_REQUEST / 400`. | Exhaustion returns canonical `OVERLOADED / 529` while preserving the redacted upstream HTTP 400 detail and request ID. | | NVIDIA-specific wording could have leaked into shared classification. | The exact marker stays NVIDIA-owned; shared provider policy owns canonical semantics, diagnostics, and scheduling, while API wire mapping remains unchanged. | | The degraded-function boundary had no deterministic coverage. | Tests cover recovery, exact exhaustion, near misses, provider isolation, raw exception preservation, trace metadata, and credential redaction; all five CI gates pass with 2,214 tests passed and 7 skipped. | | Package version was 3.5.7. | Package version is 3.5.8 with an updated lockfile. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR teaches NVIDIA NIM degraded-function failures to use the shared overload retry path. The main changes are: - Adds a provider-specific failure override hook. - Maps the exact NVCF degraded-function HTTP 400 marker to canonical overload semantics. - Threads the override through shared retry and final failure classification. - Adds focused tests for retry, exhaustion, near misses, provider isolation, redaction, and trace metadata. - Updates docs, smoke coverage metadata, package version, and lockfile. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge after considering the narrow NVIDIA error-body match. The retry and final classification flow preserve the raw upstream error, and non-NVIDIA providers keep the existing 400 behavior. src/free_claude_code/providers/nvidia_nim/client.py should harden the degraded marker lookup for nested SDK error bodies. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex produced a proof for the posted P1 finding and linked it to the corresponding review comment. - A contract-validation proof was captured showing the full command transcript with timestamps, working directory, commands, pytest verbose output, and exit codes. - The focused Nvidia degraded retry test ran and completed with 8 passed in 2.43s. - A broader test run reported worker crashes, with 3 failed and 70 passed. - A serial test for test\_openai\_compat\_5xx\_retry.py timed out after a hang, exiting at 124. <a href="https://app.greptile.com/trex/runs/14112472/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/nvidia_nim/client.py | Adds NVIDIA-specific degraded-function detection; the match is narrow and only reads a top-level `detail` field. | | src/free_claude_code/providers/rate_limit.py | Uses provider overrides for retry qualification while preserving the raw exception after retry exhaustion. | | src/free_claude_code/providers/failure_policy.py | Adds the shared override type and final classification hook, plus a reusable overloaded-provider failure constructor. | | src/free_claude_code/providers/transports/openai_chat/transport.py | Passes provider-specific failure overrides into stream creation retries and final failure mapping. | | tests/providers/test_nvidia_nim_degraded_retry.py | Adds coverage for NVIDIA degraded-function retry success, retry exhaustion, non-matching 400s, provider isolation, redaction, and traces. | </details> <!-- greptile_failed_comments --> <h3>Comments Outside Diff (1)</h3> 1. General comment <a href="#"><img alt="P1" src="https://greptile-static-assets.s3.amazonaws.com/badges/p1.svg?v=9" align="top"></a> **Adjacent OpenAI-compatible 5xx exhaustion test crashes workers and hangs serially** - **Bug** - The requested adjacent retry/rate-limit coverage is not stable on current head. Under xdist, three parameters of `test_nim_stream_openai_5xx_exhausted_emits_user_message` caused workers to terminate unexpectedly, producing an exit code 1. A follow-up serial rerun passed the 500 parameter but timed out after 60 seconds while running the 502 parameter, indicating the exhausted 5xx path can hang or terminate the worker rather than completing deterministically. - **Cause** - The executed evidence points to the OpenAI-compatible NIM 5xx exhaustion path exercised by `tests/providers/test_openai_compat_5xx_retry.py::test_nim_stream_openai_5xx_exhausted_emits_user_message`. The exact root cause was not isolated within the validation budget, but the affected area overlaps changed retry/transport handling files. - **Fix** - Investigate the OpenAI-compatible streaming 5xx exhaustion path for non-terminating retry/stream cleanup or worker-fatal behavior. Ensure exhausted 5xx cases raise the expected user-facing error promptly for all status parameters, then rerun the adjacent command and the targeted serial parametrized test. <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> <!-- /greptile_failed_comments --> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22ali%2Fretry-nim-degraded-function%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22ali%2Fretry-nim-degraded-function%22.%0A%0AFix%20the%20following%201%20code%20review%20issue.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%201%0Asrc%2Ffree_claude_code%2Fproviders%2Fnvidia_nim%2Fclient.py%3A126-127%0A**Nested%20Error%20Body%20Misses%20Retry**%0A%0AWhen%20the%20OpenAI%20SDK%20exposes%20a%20400%20response%20as%20the%20common%20nested%20error%20shape%2C%20such%20as%20%60%7B%22error%22%3A%20%7B%22message%22%3A%20%22Function%20id%20...%3A%20DEGRADED%20function%20cannot%20be%20invoked%22%7D%7D%60%2C%20this%20lookup%20never%20sees%20the%20degraded%20marker.%20That%20path%20falls%20back%20to%20the%20shared%20400%20handling%2C%20so%20a%20transient%20degraded%20NVCF%20function%20still%20fails%20after%20one%20attempt%20instead%20of%20using%20the%20provider%20retry%20budget.%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1064&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (1): Last reviewed commit: ["Retry degraded NVIDIA NIM functions as o..."](https://github.com/alishahryar1/free-claude-code/commit/2cc7f056a73d1baefb1e1395b181d26efc32dc9e) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43555691)</sub> > Greptile also left **1 inline comment** on this PR. <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
5ffa47fbc3 |
Make response stream lifetimes explicit (#1060)
## Problem Client disconnects and response-start send failures could abandon a prefetched provider stream and its generation lease. Re-yielding iterators and response-proxy middleware left no owner that closed the complete body chain before runtime release. ## Changes | Before | After | | --- | --- | | Starlette body iteration indirectly owned stream cleanup and lease release. | One FCC streaming response surrounds the real ASGI send, closes the body transitively, then releases the lease exactly once. | | The prefetched first-frame generator could not close its tail before replay began. | An explicit closeable replay iterator owns the prefetched tail in every commit state. | | Tracing, execution, Responses conversion, and native transport transforms re-yielded inputs without closing them. | Every retained transform closes its direct input; redundant transport wrappers are removed while provider construction failures remain deferred. | | Function-style correlation middleware proxied and canceled streaming responses. | Pure ASGI correlation spans the complete stream, preserves request headers and log context, and keeps the catch-all 500 fallback correlated. | | Repeated cancellation could interrupt pre-start and post-start cleanup. | Shielded completion tasks finish body closure before release and then restore caller cancellation. | | The package version was 3.5.5. | The package version is 3.5.6; full CI passes with 2,162 tests and stable live API/provider/disconnect/client smoke passes 63 scenarios. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR makes streaming response ownership explicit across the API path. The main changes are: - Adds a managed streaming response that closes the body chain before releasing provider resources. - Adds a prefetched replay iterator for first-frame commit handling. - Moves request correlation to pure ASGI middleware for full-stream context. - Propagates direct-input closure through execution, tracing, Responses conversion, and provider transports. - Bumps the package version and updates tests for stream cleanup behavior. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Validated the execution environment by reviewing the environment proof log, confirming uv 0.11.28, CPython 3.14.0, a repo-local virtual environment, and exit code 0. - Verified that the requested test command was executed, based on the test proof log. - Confirmed the test run completed successfully with 91 tests passing in 3.63 seconds, as shown in the test proof log. <a href="https://app.greptile.com/trex/runs/14099258/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/api/response_streams.py | Adds the managed response owner, first-frame replay iterator, and shielded cleanup flow. | | src/free_claude_code/api/request_ids.py | Adds pure ASGI request correlation and response-start header injection. | | src/free_claude_code/core/trace.py | Adds shared stream input closure tracing and closes traced inputs on exit. | | src/free_claude_code/application/execution.py | Closes provider stream iterators from the executor wrapper when streaming ends. | | src/free_claude_code/providers/transports/anthropic_messages/transport.py | Returns provider runner streams directly and closes layered SSE iterators explicitly. | </details> <sub>Reviews (2): Last reviewed commit: ["Make response stream lifetimes explicit"](https://github.com/alishahryar1/free-claude-code/commit/cb698c62c08924d5f80a1cea7dbd19c0b8af26a2) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43527620)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
5b35ac7c49 |
Make provider stream ownership exact and typed (#1059)
## Problem Provider stream runners lived in sibling modules, accepted `transport: Any`, and reached through private transport state. The split hid ownership behind an import-cycle workaround and prevented type checking from protecting the collaboration. ## Changes | Before | After | | --- | --- | | OpenAI-chat and native Messages streaming and recovery were split across six modules. | Each transport module owns one exactly typed private request runner and its recovery lifecycle; four obsolete modules are deleted without shims. | | Native streaming retained an unused mode export and transformation hooks. | Native streaming directly owns the single normalization path used by llama.cpp and Ollama. | | Provider retry, holdback, continuation, tool salvage, tracing, and subclass hooks crossed an untyped backchannel. | Existing behavior and provider-specific hooks remain unchanged behind checked `OpenAIChatTransport` and `AnthropicMessagesTransport` collaborators. | | Architecture checks did not cover this internal ownership boundary. | A generic AST contract rejects untyped transport collaborators and cross-module private transport access. | | Package version was `3.5.4`. | Package version is `3.5.5`; full CI passes with 2,136 tests and configured-provider behavior was exercised live. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves provider streaming ownership into typed transport modules. The main changes are: - OpenAI-chat stream and recovery logic now lives in `openai_chat/transport.py`. - Anthropic Messages stream and recovery logic now lives in `anthropic_messages/transport.py`. - Obsolete stream and recovery helper modules were removed. - Import-boundary tests now check typed transport collaborators and private transport access. - The package version and architecture notes were updated. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the provider contract tests; the test suite completed with 19 passing tests and EXIT\_CODE 0. - Initiated the provider regression test run, which was terminated before completion with EXIT\_CODE 137. - Executed Ruff provider checks; all checks passed with EXIT\_CODE 0. - Executed Ty provider type checks; all checks passed with EXIT\_CODE 0. <a href="https://app.greptile.com/trex/runs/14096487/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/transports/anthropic_messages/transport.py | Native streaming and recovery were moved into a typed per-request runner. | | src/free_claude_code/providers/transports/openai_chat/transport.py | OpenAI-chat streaming and recovery were moved into a typed per-request runner. | | tests/contracts/test_import_boundaries.py | The transport boundary contract now rejects nested `Any` transport annotations and private transport access from helper modules. | </details> <sub>Reviews (2): Last reviewed commit: ["Make provider stream owners exact and ty..."](https://github.com/alishahryar1/free-claude-code/commit/30d0c88e7a28e4c7ac44cc622f52deca5cf83b93) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43519560)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
23ed6bc87e |
Make provider backoff admission monotonic (#1053)
## Problem Concurrent provider requests could miss or shorten a reactive backoff while waiting for proactive admission. Separate gate commits could also waste quota or release requests as an expiry burst. ## Changes | Before | After | | --- | --- | | Reactive deadlines were checked once and replaced by later backoffs. | Reactive deadlines are rechecked until clear and only extended by later backoffs. | | Proactive capacity was recorded before the final reactive decision. | Conditional commit records proactive capacity only while the reactive deadline is clear. | | The limiter exposed an ambiguous `set_blocked` callback. | Both transport families use the explicit `extend_reactive_block` operation. | | Cross-gate races and wait cancellation lacked deterministic coverage. | Deterministic tests cover conditional commit, extensions, non-shortening, and cancellation. | | The architecture left gate interaction implicit. | The architecture defines monotonic two-gate admission without wasted quota or expiry bursts. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR makes provider backoff admission monotonic. The main changes are: - Conditional proactive admission before recording quota. - Reactive waits that recheck extended deadlines. - Monotonic reactive block extension for retry and stream failure paths. - Tests for admission retries, deadline extension, non-shortening, and cancellation. - Version and architecture updates for the new limiter behavior. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the provider-rate-limit pytest suite; all 53 tests completed and passed in 4.02 seconds with EXIT\_CODE: 0. - Started the transport-backoff pytest run; it advanced through most tests before xdist workers were terminated and the process exited with EXIT\_CODE: 143. - Verified that no live provider credentials, Docker containers, or external service smoke tests were used. <a href="https://app.greptile.com/trex/runs/14088781/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/core/rate_limit.py | Adds conditional sliding-window admission without recording quota on rejection. | | src/free_claude_code/providers/rate_limit.py | Reworks provider admission to combine reactive waits with final proactive commit checks. | | src/free_claude_code/providers/transports/anthropic_messages/stream.py | Routes stream rate-limit marking through monotonic reactive block extension. | | src/free_claude_code/providers/transports/openai_chat/stream.py | Routes stream rate-limit marking through monotonic reactive block extension. | | tests/core/test_strict_sliding_window.py | Covers rejected conditional admission and commit-time timestamp recording. | | tests/providers/test_provider_rate_limit.py | Covers reactive admission retries, extended deadlines, cancellation, and non-shortening behavior. | </details> <sub>Reviews (2): Last reviewed commit: ["Make provider backoff admission monotoni..."](https://github.com/alishahryar1/free-claude-code/commit/a5c637393d1996e51c195875636ecb0536522932) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43504969)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
d428b5904a |
Replace historical architecture tests with declarative boundaries (#1050)
## Problem Architecture contracts mixed real dependency rules with deleted-module tombstones and exact internal file inventories. Correct refactors therefore had to preserve history instead of the current ownership model. ## Changes | Before | After | | --- | --- | | Cross-package rules were duplicated across narrow source and layout assertions. | One least-privilege matrix and AST scanner enforce every production package edge. | | Bare imports, undeclared ownership roots, and module cycles could escape generic enforcement. | Namespaced imports, initialized owners, exact exceptions, and an acyclic module graph are enforced. | | Responses, messaging-tree, and optional dependency ownership relied on scattered checks. | Facade use and lazy optional dependency owners are declared and verified centrally. | | The messaging facade re-exported workflow, persistence, and parsing internals. | The messaging facade exposes only ingress values, platform ports, and managed-session protocols. | | Migration-era tests froze deleted modules and internal filenames. | Customer contracts remain while obsolete tombstones and layout inventories are removed. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR replaces historical architecture checks with declarative import-boundary rules. The main changes are: - Added a documented package dependency matrix and facade ownership rules. - Narrowed the messaging package facade to its supported extension surface. - Moved messaging tree consumers to the `messaging.trees` facade. - Reworked architecture tests around AST scanning, optional dependency owners, and acyclic imports. - Bumped the package version and refreshed the lockfile. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The architecture contracts tests were executed as part of the general contract validation. - The test run completed with exit code 0, indicating success. - The pytest summary shows 26 tests passed in 3.99 seconds. - An artifact log captures the exact command, working directory, environment recreation output, and verbose test item names for audit. <a href="https://app.greptile.com/trex/runs/14085499/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/messaging/__init__.py | Narrows the top-level messaging exports to the documented supported extension types. | | tests/contracts/test_import_boundaries.py | Replaces historical layout checks with declarative import-boundary enforcement. | | ARCHITECTURE.md | Documents the current dependency matrix, facade boundaries, and optional dependency owners. | | src/free_claude_code/messaging/node_event_pipeline.py | Uses the messaging tree facade for `NodeClaim`. | | src/free_claude_code/messaging/node_runner.py | Uses the messaging tree facade for queue, snapshot, cancellation, and claim types. | </details> <sub>Reviews (2): Last reviewed commit: ["Enforce architecture with declarative im..."](https://github.com/alishahryar1/free-claude-code/commit/68508f3d2bad19d4d1b8a9ca5d181e1433db4d71) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43498975)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
701a394697 |
Make provider capabilities semantic and typed (#1047)
## Problem Provider metadata mixed descriptive strings and transport-family names with product policy. The Messages API had to know provider implementation details to decide server-tool behavior, while locality and thinking metadata had ambiguous owners. ## Changes | Before | After | | --- | --- | | Provider descriptors exposed transport names and untyped capability strings. | Provider descriptors expose an immutable semantic value containing only locality and server-tool passthrough. | | The Messages handler derived provider sets from transport families. | `ResolvedModel` carries catalog capabilities and the handler asks the resolved semantic policy. | | Discovery, Admin, and smoke selection inferred locality from credentials or string membership. | Discovery, Admin, and smoke selection read the catalog-owned `local` capability. | | Provider-wide thinking strings competed with discovered model metadata. | `ProviderModelInfo.supports_thinking` remains the sole per-model thinking authority. | | Invalid mapped providers could become raw catalog lookup failures during the migration. | Routing and provider construction share the canonical typed unknown-provider error. | | LM Studio smoke contracts described a native Anthropic transport. | LM Studio smoke contracts describe its OpenAI-chat-backed Messages path. | | Tests encoded transport identities as product behavior. | Tests prove both directions of capability-versus-identity independence while preserving existing wire behavior and diagnostics. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves provider policy from transport names to typed catalog capabilities. The main changes are: - Added immutable provider capability values for locality and server-tool passthrough. - Routed provider capabilities through `ResolvedModel` for Messages API policy. - Updated admin status, discovery, smoke config, docs, and tests to read semantic capabilities. - Centralized unknown-provider diagnostics for routing and provider construction. - Bumped the package version and lockfile to `3.4.22`. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The Pytest run for the provider capabilities changes was executed and completed successfully, as shown by the log's EXIT\_CODE: 0. - The Ruff linter run for the provider capabilities changes was executed and completed successfully, as shown by the log's EXIT\_CODE: 0. - The focused suite exercised the changed code paths via repository tests rather than relying solely on static inspection. <a href="https://app.greptile.com/trex/runs/14074560/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/config/provider_catalog.py | Replaces transport and string capability metadata with typed immutable provider capabilities. | | src/free_claude_code/application/routing.py | Adds catalog capabilities to resolved models and uses the shared unknown-provider error. | | src/free_claude_code/api/handlers/messages.py | Uses resolved semantic capabilities to decide server-tool passthrough. | | src/free_claude_code/config/admin/status.py | Classifies local providers through the new catalog capability. | | src/free_claude_code/providers/runtime/discovery.py | Uses typed locality to decide which providers are discovered only when referenced. | | src/free_claude_code/application/errors.py | Adds a shared constructor for canonical unknown-provider messages. | </details> <sub>Reviews (1): Last reviewed commit: ["Make provider capabilities semantic and ..."](https://github.com/alishahryar1/free-claude-code/commit/b2732ef74d61c632ccae332a19aaa655f45a96e7) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43475337)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
e22a38b2c2 |
Canonicalize provider failure and retry ownership (#1046)
## Problem Provider SDK classification, retry policy, canonical failures, and downstream wire errors shared exception types across layers. That blurred ownership and let cleanup or provisional Responses tool failures mask the real provider diagnostic. ## Changes | Before | After | | --- | --- | | Provider failures carried Anthropic wire types and core code classified OpenAI/httpx errors. | Protocol-neutral `ExecutionFailure` values cross layers, providers classify SDK errors, and protocol packages map wire types. | | Provider adapters could author terminal wire events. | The HTTP commit boundary selects non-2xx JSON or a protocol terminal event with one ingress request ID. | | Retry policy and diagnostic handling were spread across core and provider modules. | Providers own the unchanged retry budgets while neutral core utilities own bounded credential redaction. | | Stream cleanup could replace an already-mapped provider failure. | Cleanup records safe metadata and preserves the canonical failure, status, and diagnostic. | | An incomplete Responses tool could preempt a later provider failure. | Tool-finalization errors remain provisional so canonical provider failures take precedence. | | Readiness failures reused provider exception types. | Application-owned errors represent deterministic validation and availability phases without terminal retry headers. | | Legacy exception and recovery owners remained importable. | Obsolete modules are deleted without shims, architecture rules enforce the boundaries, and package version is 3.4.21. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR canonicalizes provider failure handling across the API boundary. The main changes are: - Adds protocol-neutral execution failure values and safe diagnostics. - Moves SDK and HTTP failure classification into provider-owned policy. - Lets Messages and Responses choose their own wire error payloads. - Preserves canonical failures across stream cleanup and committed stream failures. - Makes incomplete Responses tool errors provisional until finalization. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. No files need attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the API failure contract suite and related tests (tests/api/test\_execution\_failure\_contract.py, tests/core/test\_failure\_protocol\_mapping.py, tests/providers/test\_execution\_failure\_boundary.py, tests/providers/test\_failure\_policy.py); 48 passed in 3.36s. - Ran the streaming boundaries tests including response streams, stream recovery, and streaming errors; 70 passed in 5.36s. - Ran the OpenAI responses tests; 20 passed in 4.42s. <a href="https://app.greptile.com/trex/runs/14071383/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/providers/transports/http.py | Adds cleanup-safe stream closing that preserves established outcomes. | | src/free_claude_code/core/openai_responses/stream.py | Preserves canonical execution failures when committed Responses streams fail. | | src/free_claude_code/core/openai_responses/streaming/assembler.py | Keeps malformed tool-call errors provisional so later provider failures can win. | | src/free_claude_code/core/failures.py | Defines neutral failure kinds and exception-group lookup for execution failures. | </details> <sub>Reviews (2): Last reviewed commit: ["Preserve canonical outcomes in grouped a..."](https://github.com/alishahryar1/free-claude-code/commit/f57f21241dbe582985627ed4fb40734b2c656809) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43468297)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
f8c21a48f2 |
Introduce a typed application boundary for provider execution (#1045)
## Problem The HTTP adapter owned model routing, provider execution, and runtime-facing contracts, so API handlers depended on provider implementation types. Provider preflight also discovered private request builders dynamically, obscuring the boundary that must fail before streaming begins. ## Changes | Before | After | | --- | --- | | `api/` owned model routing and shared provider execution. | `application/` owns routing and a settings-independent `ProviderExecutor`. | | API handlers accepted `BaseProvider` callbacks. | API handlers consume the narrow structural `ProviderPort`. | | `BaseProvider` discovered `_build_request_body` dynamically. | Both transport families implement explicit abstract preflight, with LM Studio composing context validation. | | Request leases, task control, and provider model metadata had adapter/provider owners. | Application-owned ports and immutable values define those cross-package contracts. | | Boundary direction was implicit. | Architecture contracts and documentation enforce the final dependency direction. | | Package version was `3.4.19`. | Package version is `3.4.20`, with the lockfile updated. | | Coverage followed the old module layout. | Deterministic boundary/preflight regressions and live Messages/Responses smokes cover the new shape. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds a typed application boundary for provider execution. The main changes are: - New `application` package for routing, execution, ports, and model metadata. - API handlers now call application-owned routing and provider execution. - Provider preflight is now explicit on the transport families. - Runtime API composition now uses a task-control port for `/stop`. - Import-boundary tests, smoke references, docs, version, and lockfile were updated. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. None. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran a deterministic pytest run for the provider boundary preflight, which completed with exit code 0 and 102 tests passing in 4.77 seconds. - Launched the environment presence check as part of the preflight, which completed with exit code 0 and confirmed OPENCODE\_API\_KEY=\[REDACTED\] matched. - Attempted the live provider smoke test, which completed with exit code 0 and 2 tests skipped due to incomplete smoke configuration. <a href="https://app.greptile.com/trex/runs/14067905/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/application/execution.py | Moves shared provider execution into the application layer and keeps eager preflight before token counting and streaming. | | src/free_claude_code/application/ports.py | Adds structural provider, request-runtime, and task-control protocols used across the new boundary. | | src/free_claude_code/api/routes.py | Updates route composition to use the application provider resolver and task-control stop path. | | src/free_claude_code/providers/base.py | Makes provider preflight explicit by requiring subclasses or transport bases to implement it. | | src/free_claude_code/providers/transports/openai_chat/transport.py | Adds OpenAI-chat preflight through the same request-body builder used by streaming. | | src/free_claude_code/providers/transports/anthropic_messages/transport.py | Adds native Messages preflight through the native request-body builder. | | src/free_claude_code/providers/model_listing.py | Keeps provider model-list parsing while moving `ProviderModelInfo` ownership to the application layer. | | src/free_claude_code/runtime/bootstrap.py | Passes the runtime object through the new `tasks` service slot. | | tests/contracts/test_import_boundaries.py | Extends import-boundary tests for the new application package. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart LR API[api handlers and routes] --> Routing[application.routing] API --> Executor[application.execution] API --> Ports[application.ports] Executor --> ProviderPort[ProviderPort] ProviderPort --> Preflight[preflight_stream] ProviderPort --> Stream[stream_response] Runtime[runtime bootstrap and provider manager] --> Ports Providers[providers] --> Metadata[application.model_metadata] Executor --> Core[core anthropic and trace] Routing --> Config[config settings and model refs] ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart LR API[api handlers and routes] --> Routing[application.routing] API --> Executor[application.execution] API --> Ports[application.ports] Executor --> ProviderPort[ProviderPort] ProviderPort --> Preflight[preflight_stream] ProviderPort --> Stream[stream_response] Runtime[runtime bootstrap and provider manager] --> Ports Providers[providers] --> Metadata[application.model_metadata] Executor --> Core[core anthropic and trace] Routing --> Config[config settings and model refs] ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Introduce typed application boundary"](https://github.com/alishahryar1/free-claude-code/commit/4cdcf97231c812c6f568ca3d74af7ce759d7f2dc) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43462788)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
4a0a0360de |
Move protocol models to their protocol owners (#1044)
## Problem Anthropic Messages and OpenAI Responses wire models lived under the inbound API adapter. Neutral protocol and provider code therefore duck-typed requests, obscuring ownership and weakening dependency boundaries. ## Changes | Before | After | | --- | --- | | The API package owned Anthropic and Responses protocol models. | Each protocol package owns and publicly exports its wire models. | | Core and provider request paths accepted `Any` and probed known fields with `getattr()`. | Core, transports, and providers consume concrete `MessagesRequest` values. | | Responses conversion and streaming received a dumped request mapping. | Responses conversion and streaming receive one concrete `OpenAIResponsesRequest`. | | Anthropic request snapshots lived in generic tracing code. | Anthropic request snapshots live with the protocol while generic tracing stays protocol-independent. | | Protocol tests and provider request doubles reflected the old API ownership. | Protocol tests live under core and provider tests construct real wire requests. | | The API model package mixed protocol and model-catalog schemas. | The API model package is removed, with catalog schemas beside catalog construction and no compatibility shim. | | Package version was `3.4.18`. | Package version is `3.4.19` with an updated lockfile. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves protocol request models to their protocol-owned packages. The main changes are: - Anthropic Messages models now live under `core.anthropic`. - OpenAI Responses models now live under `core.openai_responses`. - API handlers, routes, providers, and tests now use concrete protocol request types. - Anthropic request snapshots moved beside the Anthropic protocol models. - API model catalog schemas were kept with catalog response construction. - The package version and lockfile were updated. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues were found in the changed code. Internal callers were updated to pass the new concrete protocol models, and no stale internal imports from the removed API model package were identified. No files need attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The Pytest run for protocol ownership focused tests completed, showing 71 passed in 6.24s and EXIT\_CODE: 0. - A protocol import smoke script was generated for the import/conversion/trace workflow. - The protocol import smoke run completed successfully, including model ownership output, adapter payload evidence, and trace snapshot evidence, with EXIT\_CODE: 0. <a href="https://app.greptile.com/trex/runs/14066157/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/core/anthropic/models.py | Anthropic wire request and response models moved under the Anthropic protocol package. | | src/free_claude_code/core/anthropic/native_messages_request.py | Native Anthropic serialization now expects concrete `MessagesRequest` instances. | | src/free_claude_code/core/anthropic/conversion.py | OpenAI chat conversion now reads fields directly from `MessagesRequest`. | | src/free_claude_code/core/anthropic/request_snapshot.py | Anthropic request snapshotting moved from generic tracing into the protocol package. | | src/free_claude_code/core/openai_responses/models.py | OpenAI Responses ingress models moved under the Responses protocol package. | | src/free_claude_code/core/openai_responses/input.py | Responses conversion now consumes the concrete request model instead of a dumped mapping. | | src/free_claude_code/core/openai_responses/streaming/assembler.py | Responses stream assembly now reads request attributes from `OpenAIResponsesRequest`. | | src/free_claude_code/api/routes.py | Routes now import protocol request models from their new core owners. | | src/free_claude_code/api/model_catalog.py | Model-list response schemas now live with model catalog construction. | </details> <details open><summary><h3>Flowchart</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% flowchart LR API[API routes and handlers] --> Anthropic[core.anthropic models and helpers] API --> Responses[core.openai_responses models and adapter] Responses --> Anthropic Providers[Provider clients and transports] --> Anthropic Anthropic --> Trace[core.trace sanitization] API --> Catalog[api.model_catalog response schemas] ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% flowchart LR API[API routes and handlers] --> Anthropic[core.anthropic models and helpers] API --> Responses[core.openai_responses models and adapter] Responses --> Anthropic Providers[Provider clients and transports] --> Anthropic Anthropic --> Trace[core.trace sanitization] API --> Catalog[api.model_catalog response schemas] ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Move protocol models to their protocol o..."](https://github.com/alishahryar1/free-claude-code/commit/f1be5c1af4a10da710f80b5f9e7f6044a601f6af) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43459428)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
4951983b5e |
Replace global runtime resources with explicit ownership (#1042)
## Problem Provider, messaging, and transcription resources relied on process-global state, leaving replacement, cancellation, and shutdown ownership ambiguous. Separate server lifetimes could share event-loop-bound resources or retain failed cleanup work. ## Changes | Before | After | | --- | --- | | Provider clients found limiters through global singleton and scoped registries. | Each provider instance receives and owns one explicitly constructed limiter. | | Messaging queues and voice pipelines relied on singleton or module-global state. | Each platform owns its limiter and outbox, while the application owns one injected transcriber. | | Messaging shutdown mixed ingress, active work, delivery, and SDK cleanup. | Application shutdown quiesces ingress, drains work, closes delivery, then releases transcription and providers. | | Cancelled or failed provider cleanup could be forgotten or treated as complete. | The provider manager retains shielded generation and unpublished-runtime cleanup until it succeeds. | | Discord and Telegram startup tasks could outlive or poison runtime readiness. | Platform runtimes observe long-lived tasks and retry only independently repeatable lifecycle steps. | | Constructor-captured security and diagnostic settings appeared hot-applicable. | Admin marks those settings restart-required so applied policy matches the running resource graph. | | Lifecycle races lacked direct ownership coverage. | Deterministic cancellation, retry, isolation, teardown, and live smoke contracts protect the final ownership model. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves runtime resources from global state into explicitly owned application objects. The main changes are: - Provider generations own their rate limiters and cleanup tasks. - Messaging platforms own their limiter, outbox, ingress, and delivery lifecycle. - Application shutdown now runs through ordered cleanup gates. - Voice transcription is injected as an owned runtime resource. - Admin config marks constructor-captured settings as restart-required. </details> <h3>Confidence Score: 4/5</h3> The shutdown path needs a bounded cleanup result before merging. Cleanup steps that hang never reach the retryable incomplete-shutdown path. ASGI shutdown can remain stuck while waiting for an external SDK, transcriber, workflow, or provider cleanup. The retry ownership model works only after cleanup returns or raises. src/free_claude_code/runtime/application.py <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - T-Rex ran the requested verification, but its local artifact references were not uploaded. - The validation run completed successfully with EXIT\_CODE: 0 and 62 tests passed in 3.91 seconds, using the command uv run pytest -vv tests/runtime/test\_application\_runtime.py tests/runtime/test\_provider\_manager.py tests/providers/test\_provider\_runtime.py. <a href="https://app.greptile.com/trex/runs/14064214/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/runtime/application.py | Refactors shutdown into ordered retryable cleanup gates, but cleanup awaitables can still block shutdown forever. | | src/free_claude_code/runtime/asgi.py | Reports incomplete runtime shutdown when `close()` returns false. | | src/free_claude_code/runtime/provider_manager.py | Adds owned provider cleanup retry state and shielded generation cleanup. | </details> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22refactor%2Fruntime-owned-resources%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22refactor%2Fruntime-owned-resources%22.%0A%0AFix%20the%20following%201%20code%20review%20issue.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%201%0Asrc%2Ffree_claude_code%2Fruntime%2Fapplication.py%3A59%0A**Cleanup%20Await%20Blocks%20Shutdown**%0A%0AWhen%20a%20platform%20SDK%20stop%2C%20workflow%20drain%2C%20transcriber%20close%2C%20or%20provider%20cleanup%20hangs%2C%20this%20helper%20waits%20forever%20and%20never%20returns%20%60False%60.%20ASGI%20shutdown%20stays%20stuck%20in%20%60runtime.close%28%29%60%20instead%20of%20reporting%20an%20incomplete%20shutdown%2C%20so%20the%20retained%20resource%20graph%20cannot%20be%20retried%20cleanly.%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1042&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (2): Last reviewed commit: ["Report incomplete runtime shutdown to AS..."](https://github.com/alishahryar1/free-claude-code/commit/338b2bd179c3875b15bbd52818dd04c780e5d46d) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43454593)</sub> > Greptile also left **1 inline comment** on this PR. **Context used:** - Context used - CLAUDE.md ([source](https://app.greptile.com/alishahryar1/github/Alishahryar1/free-claude-code/-/custom-context?memory=d2fd24d8-0dec-4faf-8ee4-e085e215a2f8)) <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
160d63370b |
Establish single-owner runtime with stream-safe provider hot swaps (#1036)
## Problem Provider runtime ownership was split between lifecycle code and mutable FastAPI state, so Admin replacements could leak the new runtime, double-close the old runtime, or close providers still serving active streams. The API package also owned concrete process composition, obscuring subsystem boundaries. ## Changes | Before | After | | --- | --- | | FastAPI routes inspected several concrete `app.state` resources. | FastAPI receives one explicit `ApiServices` boundary and stores only `app.state.services`. | | Admin Apply persisted config and directly replaced one runtime reference. | Admin Apply validates a candidate, commits atomically, and publishes it through the single runtime owner. | | Provider replacement could close clients used by active streams. | Generation leases retain old providers until each streaming or non-streaming response finishes. | | Provider generations owned discovery state and model metadata. | `ProviderRuntimeManager` owns one application-lifetime catalog and one discovery task across replacements. | | API modules composed provider, messaging, and managed CLI resources. | `runtime.bootstrap` composes concrete subsystems and `ApplicationRuntime` owns their lifecycle. | | Admin config, server URLs, and gateway model IDs lived under the API package. | Admin config lives under `config`, server URLs live under `config`, and gateway IDs live under `core`. | | Messaging restoration and shutdown persistence were coordinated externally. | `MessagingWorkflow` owns snapshot restoration and final persistence flushing. | | Hot-swap behavior lacked a real process-level race scenario. | Deterministic ownership tests and a credential-free subprocess smoke hold provider A while new requests switch to provider B. | | Package version was `3.4.16`. | Package version is `3.4.17` with an updated lockfile. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR centralizes server runtime ownership and provider hot swaps. The main changes are: - Adds `ApplicationRuntime` and `ProviderRuntimeManager` as the process owners. - Moves FastAPI to an explicit `ApiServices` boundary. - Retains provider generations until request and stream responses finish. - Moves admin config, server URL, and gateway model ID modules to neutral package owners. - Updates admin apply to validate, persist, and publish provider-only changes through the runtime owner. - Adds runtime ownership tests and a credential-free smoke scenario. </details> <h3>Confidence Score: 5/5</h3> This looks safe to merge. No blocking issues found in the changed code. The provider lease path releases resources on normal completion, stream close, and cancellation. The admin apply path keeps restart-required changes separate from provider-only hot swaps, and repository import paths appear updated for the moved modules. No files need follow-up attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the T-Rex smoke test command and confirmed it completed with exit code 0 and pytest passing. - Monitored runtime ownership activity during the run, including a provider A stream request on model-a generation 1, an admin publish to generation 2, a new request on model-b generation 2, and the completion of the original generation 1 stream. - Collected smoke result artifacts from the .smoke-results area for gateway 1, gateway 0, and main, and made them available for review. - Opened the smoke report JSON artifacts to review the summarized outcomes for each target environment. <a href="https://app.greptile.com/trex/runs/14008910/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/runtime/provider_manager.py | Adds provider generation ownership, request leases, replacement, discovery refresh, and shutdown cleanup. | | src/free_claude_code/runtime/application.py | Adds the process-level owner for startup, shutdown, admin operations, messaging, and session control. | | src/free_claude_code/api/routes.py | Routes now acquire provider generation leases and bind them to response lifetime. | | src/free_claude_code/api/response_streams.py | Adds response lifetime binding so retained resources release after stream completion, cancellation, or close. | | src/free_claude_code/config/admin/persistence.py | Moves admin config persistence into `config` and adds prepared validation plus atomic managed-env commits. | | src/free_claude_code/runtime/bootstrap.py | Adds the production composition root for logging, runtime owners, services, and ASGI wiring. | | src/free_claude_code/api/__init__.py | Removes package-level API re-exports as part of the HTTP adapter boundary cleanup. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant Client participant API as FastAPI Route participant Manager as ProviderRuntimeManager participant Lease as Generation Lease participant Runtime as Provider Generation participant Admin as Admin Apply Client->>API: Request /v1/messages or /v1/responses API->>Manager: acquire() Manager-->>API: lease for current generation API->>Lease: resolve_provider() Lease->>Runtime: use provider instance Runtime-->>Client: response body or stream Admin->>Manager: replace(candidate settings) Manager->>Manager: publish new generation Manager->>Manager: retire old generation Client-->>API: response completes or disconnects API->>Lease: release() Lease->>Manager: decrement active leases Manager->>Runtime: cleanup retired generation when drained ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant Client participant API as FastAPI Route participant Manager as ProviderRuntimeManager participant Lease as Generation Lease participant Runtime as Provider Generation participant Admin as Admin Apply Client->>API: Request /v1/messages or /v1/responses API->>Manager: acquire() Manager-->>API: lease for current generation API->>Lease: resolve_provider() Lease->>Runtime: use provider instance Runtime-->>Client: response body or stream Admin->>Manager: replace(candidate settings) Manager->>Manager: publish new generation Manager->>Manager: retire old generation Client-->>API: response completes or disconnects API->>Lease: release() Lease->>Manager: decrement active leases Manager->>Runtime: cleanup retired generation when drained ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["refactor: establish single-owner applica..."](https://github.com/alishahryar1/free-claude-code/commit/92fc06aa733b7acc34ad6ea50de8b6b4ce5cfac1) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43349002)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
1278d00873 |
Restore protocol-correct provider errors without client retry loops (#1033)
## Problem Provider failures before streaming were converted to HTTP 200 SSE errors, masking typed upstream statuses as malformed proxy responses. This regressed #1026's error visibility while solving client retry loops; fixes #1031. ## Changes | Before | After | | --- | --- | | Pre-start Messages failures returned HTTP 200 SSE errors. | Pre-start Messages failures return typed non-2xx JSON with `x-should-retry: false`. | | Non-streaming Messages could preserve partial content after an internal stream error. | Non-streaming Messages discard partial content and return the mapped Anthropic error. | | Post-start failures could synthesize a successful stop lifecycle. | Post-start failures emit protocol-native terminal errors without a fake success stop. | | Responses failures lacked consistent retry ownership and correlation. | Responses failures retain typed envelopes, retry suppression, response IDs, and request IDs. | | Request IDs were generated independently across layers. | One ingress-owned request ID flows through response headers, provider calls, error bodies, and traces. | | Provider exceptions owned Anthropic serialization. | Neutral Anthropic utilities own error envelopes, status mapping, and redacted diagnostics. | | Failure-path smoke expected the regressed HTTP 200 shape. | Failure-path smoke verifies typed JSON and one downstream Claude CLI request. | | Package metadata remained at 3.4.15. | Package metadata and the lockfile advance to 3.4.16. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR restores protocol-correct provider error handling for Messages and Responses. The main changes are: - Pre-start provider failures return typed JSON errors with retry suppression. - Streaming failures use protocol-native terminal events after commit. - Request IDs now flow from ingress through headers, traces, provider calls, and error bodies. - Shared Anthropic and OpenAI error helpers now shape payloads and redact diagnostics. - Provider-error smoke coverage and package metadata were updated. </details> <h3>Confidence Score: 4/5</h3> Committed streaming failure paths still expose exception-derived messages to clients. Pre-start provider error handling is more protocol-correct, and the completed-message lifecycle conflict appears fixed. Messages and Responses streams still need fixed safe messages after the HTTP response has committed. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - I executed a focused uv-run harness against the Anthropic SSE committed-stream path and reproduced a committed StreamingResponse that yielded a terminal error SSE frame containing an internal diagnostic marker. - I ran a focused uv-run against iter\_responses\_sse\_from\_anthropic and confirmed the stream was committed, with the later response.failed SSE event carrying the internal diagnostic string in response.error. - I completed provider error-handling verification, validating the deterministic pytest path and running the local /v1/responses probe script, which reported a 429 status with downstream\_call\_count: 1 and related metadata, using the exact probe to exercise the failing provider. <a href="https://app.greptile.com/trex/runs/13955393/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | src/free_claude_code/api/response_streams.py | Pre-start Messages errors now return typed JSON, but committed terminal SSE errors still use exception-derived text. | | src/free_claude_code/core/openai_responses/stream.py | Responses streams now emit protocol failure events, but committed failures still use exception-derived text. | | src/free_claude_code/core/anthropic/streaming/ledger.py | Completed Anthropic message streams now suppress late terminal errors after `message_stop`. | | src/free_claude_code/core/anthropic/errors.py | Shared Anthropic error payload helpers now add request IDs, status mapping, and diagnostic redaction. | </details> <a href="https://app.greptile.com/api/ide/codex?prompt=IMPORTANT%3A%20Work%20in%20the%20repository%20%22alishahryar1%2Ffree-claude-code%22%20on%20the%20existing%20branch%20%22fix%2Fprotocol-correct-provider-errors%22.%20Checkout%20that%20branch%20%E2%80%94%20do%20NOT%20create%20a%20new%20branch%20or%20open%20a%20new%20PR.%20Push%20your%20changes%20to%20%22fix%2Fprotocol-correct-provider-errors%22.%0A%0AFix%20the%20following%202%20code%20review%20issues.%20Work%20through%20them%20one%20at%20a%20time%2C%20proposing%20concise%20fixes.%0A%0A---%0A%0A%23%23%23%20Issue%201%20of%202%0Asrc%2Ffree_claude_code%2Fapi%2Fresponse_streams.py%3A115-117%0A**Committed%20Stream%20Exception%20Text**%0A%0AWhen%20an%20Anthropic%20stream%20fails%20after%20the%20first%20chunk%20has%20committed%2C%20this%20terminal%20frame%20still%20derives%20the%20public%20SSE%20message%20from%20the%20exception.%20For%20unknown%20SDK%20or%20runtime%20errors%2C%20%60get_user_facing_error_message%28%29%60%20can%20return%20sanitized%20%60str%28exc%29%60%2C%20so%20internal%20URLs%2C%20payload%20fragments%2C%20or%20unsupported%20secret%20formats%20can%20still%20reach%20the%20client.%0A%0A%23%23%23%20Issue%202%20of%202%0Asrc%2Ffree_claude_code%2Fcore%2Fopenai_responses%2Fstream.py%3A49%0A**Failed%20Response%20Exception%20Text**%0A%0AAfter%20a%20Responses%20stream%20has%20emitted%20%60response.created%60%2C%20this%20failure%20event%20still%20uses%20exception-derived%20text%20as%20the%20public%20error%20message.%20If%20an%20unexpected%20provider%20or%20SDK%20exception%20includes%20internal%20diagnostics%20or%20a%20credential%20shape%20outside%20the%20redaction%20patterns%2C%20the%20committed%20%60response.failed%60%20event%20exposes%20it%20to%20the%20client.%0A%0A&repo=alishahryar1%2Ffree-claude-code&pr=1033&platform=github"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodexDark.svg?v=6"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"><img alt="Fix All in Codex" src="https://greptile-static-assets.s3.amazonaws.com/badges/FixAllInCodex.svg?v=6"></picture></a> <sub>Reviews (2): Last reviewed commit: ["fix(streaming): ignore errors after mess..."](https://github.com/alishahryar1/free-claude-code/commit/157eb504e9c8e15f6591bd72391810610be078f7) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=43242122)</sub> > Greptile also left **2 inline comments** on this PR. **Context used:** - Context used - CLAUDE.md ([source](https://app.greptile.com/alishahryar1/github/Alishahryar1/free-claude-code/-/custom-context?memory=d2fd24d8-0dec-4faf-8ee4-e085e215a2f8)) <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
71a78a0c5a |
Move runtime packages under src namespace (#1029)
## Problem Runtime modules were published as generic top-level packages like `api`, `cli`, and `providers`. That shape is fragile for PyPI packaging and weakens explicit ownership boundaries. ## Changes | Before | After | | --- | --- | | Runtime code lived in root-level packages. | Runtime code lives under `src/free_claude_code/`. | | Console scripts targeted top-level modules. | Console scripts target namespaced modules. | | Tests and smoke helpers imported old package roots. | Tests and smoke helpers import `free_claude_code.*`. | | Packaging listed six root packages. | Packaging builds the single namespaced package. | | Contracts allowed old root package directories. | Contracts require the src namespace and reject old root imports. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves the runtime packages into the `src/free_claude_code` namespace. The main changes are: - Console scripts now point to `free_claude_code.*` entrypoints. - Runtime imports, tests, and smoke helpers now use the namespaced package. - Packaging now builds the single `src/free_claude_code` package. - Contract tests now reject old top-level runtime package roots and imports. </details> <h3>Confidence Score: 5/5</h3> This PR is safe to merge with minimal risk. The changes are a broad but mostly mechanical namespace and package-layout migration with updated packaging, tests, and contract coverage. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Reviewed the primary contract validation by examining the namespace validation log, which documents the exact commands executed, the working directory, exit codes, pytest output, wheel build output, install output, and import/entrypoint resolution. - Verified the wheel listing by inspecting the wheel listing artifact, confirming the available wheel filenames for the namespace validation. - Ran and inspected the isolated import/entrypoint validation harness saved as package-installed-import-check.py to validate import resolution and entrypoint exposure. - Captured and noted the wheel filename record in package-wheel-filename.txt to enable traceability of the observed artifact. <a href="https://app.greptile.com/trex/runs/13810533/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | pyproject.toml | Updates packaging to build the single `src/free_claude_code` package and retargets console scripts to namespaced modules. | | src/free_claude_code/config/env_template.py | Loads `.env.example` from packaged resources with a source-checkout fallback after the runtime package move. | | src/free_claude_code/cli/entrypoints.py | Updates CLI entrypoint imports to `free_claude_code.*` and continues to use the shared env template loader. | | src/free_claude_code/api/routes.py | Retargets API route dependencies and handlers to the namespaced package without changing route behavior. | | src/free_claude_code/api/app.py | Updates app factory imports to the namespaced package while preserving middleware, routers, and exception handling. | | src/free_claude_code/providers/runtime/factory.py | Updates lazy provider factory imports to `free_claude_code.providers.*` under the new package layout. | | tests/contracts/test_import_boundaries.py | Adds contract coverage requiring runtime packages to live under `src/free_claude_code` and rejecting old top-level imports. | | smoke/lib/child_process.py | Updates smoke child-process helpers to import CLI entrypoints from the namespaced package. | | README.md | Updates the project layout and extension guidance to refer to `src/free_claude_code` and importable `free_claude_code.*` modules. | | uv.lock | Reflects the package version bump associated with the runtime packaging move. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant User as User / CLI participant Script as Console script participant Pkg as free_claude_code package participant API as free_claude_code.api participant Runtime as free_claude_code.providers.runtime participant Provider as Provider adapter User->>Script: run fcc-server / free-claude-code Script->>Pkg: load free_claude_code.cli.entrypoints:serve Pkg->>API: create FastAPI app and routes API->>Runtime: resolve configured provider Runtime->>Provider: instantiate namespaced adapter Provider-->>Runtime: stream/model responses Runtime-->>API: provider result API-->>User: Anthropic/OpenAI-compatible response ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant User as User / CLI participant Script as Console script participant Pkg as free_claude_code package participant API as free_claude_code.api participant Runtime as free_claude_code.providers.runtime participant Provider as Provider adapter User->>Script: run fcc-server / free-claude-code Script->>Pkg: load free_claude_code.cli.entrypoints:serve Pkg->>API: create FastAPI app and routes API->>Runtime: resolve configured provider Runtime->>Provider: instantiate namespaced adapter Provider-->>Runtime: stream/model responses Runtime-->>API: provider result API-->>User: Anthropic/OpenAI-compatible response ``` </a> </details> <sub>Reviews (2): Last reviewed commit: ["Fix documented package import paths"](https://github.com/alishahryar1/free-claude-code/commit/bfa9f2704c45f3684da39657d5e13f3814e5d450) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=42950471)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
a5310f4a0f |
Fix pre-start stream failures returning HTTP 200 (#1026)
## Problem
Provider streams could commit HTTP 200 before an upstream-backed first
SSE frame was available. When setup or retry failed before usable stream
output, Claude and Codex saw a successful but broken stream instead of a
retryable non-200 error.
## Changes
| Before | After |
| --- | --- |
| API egress returned `StreamingResponse` before probing the provider
iterator. | API egress waits for the first chunk before committing
success headers. |
| Pre-start provider failures became synthetic SSE success streams. |
Pre-start provider failures raise typed errors and return Anthropic or
OpenAI JSON with non-200 status. |
| Post-start unexpected stream failures could truncate protocol output.
| Post-start failures emit terminal Anthropic error or Responses
`response.failed` frames where possible. |
| Provider tests expected pre-start final failures as SSE tails. |
Provider tests assert typed pre-start errors and preserve midstream and
tool-salvage behavior. |
<!-- greptile_comment -->
<details open><summary><h3>Greptile Summary</h3></summary>
This PR changes streaming responses so HTTP success is not committed
before the first protocol chunk. The main changes are:
- First-chunk gated streaming helpers for Anthropic and OpenAI Responses
egress.
- Non-200 JSON error responses for provider failures before stream
output starts.
- Terminal Anthropic `error` and Responses `response.failed` frames for
post-start interruptions.
- Provider transport updates that raise typed pre-start errors while
preserving retry, recovery, and tool salvage paths.
- Targeted API and provider tests plus a patch version and lockfile
update.
</details>
<h3>Confidence Score: 5/5</h3>
Safe to merge with low risk.
The changed paths keep provider retry and recovery ownership in
transports, gate HTTP success before the first chunk, and preserve
cancellation behavior. Tests cover the main Anthropic and OpenAI
Responses pre-start and post-start failure paths.
No files require special attention.
<details><summary><h3><a href="https://www.greptile.com/trex"><img
alt="T-Rex"
src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg"
height="20" align="absmiddle"></a> T-Rex Logs</h3></summary>
**What T-Rex did**
- Ran the pre-change stream-gating API tests against the previous commit
|
||
|
Ali Khokhar
|
745c38cbbe |
Move cloud providers to OpenAI-chat transport
Move remote cloud providers onto the OpenAI-chat transport and keep native Anthropic transport local-provider only. |
||
|
Ali Khokhar
|
dac6d4e88d |
Request streamed usage for OpenAI-chat providers (#1013)
## Problem OpenAI-compatible streaming providers could return accurate final usage, but FCC only requested it for DeepSeek and kept provider prompt tokens out of final Anthropic usage. ## Changes | Before | After | | --- | --- | | DeepSeek alone requested streamed usage. | The OpenAI-chat transport requests streamed usage for all OpenAI-compatible providers. | | Final input usage stayed on the local estimate even when providers returned `prompt_tokens`. | Final input usage uses provider `prompt_tokens` when available and falls back to the estimate when absent. | | Providers that rejected `stream_options.include_usage` failed the request. | Providers that reject optional usage metadata retry once without it. | | DeepSeek owned duplicated usage extraction logic. | OpenAI-chat usage extraction is shared, while DeepSeek keeps only cache-token mapping. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves streamed usage handling into the shared OpenAI-chat transport. The main changes are: - Requests `stream_options.include_usage` for OpenAI-compatible streaming providers. - Uses provider `prompt_tokens` and `completion_tokens` when streamed usage is returned. - Falls back to local token estimates when usage metadata is absent or rejected. - Retries once without `include_usage` when an upstream provider rejects the optional field. - Keeps DeepSeek-specific cache token mapping in the DeepSeek provider. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with low risk. The changed logic is localized to the OpenAI-chat streaming path and includes focused regression coverage for the new usage and retry behavior. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The team executed the general contract validation for the non-UI provider transport using a mocked chat-completion streaming surface. - Runtime artifacts, including the runtime log and before/after comparison artifacts, were produced for inspection. - The after-run summary shows 10 passed in 5.38 seconds with EXIT\_CODE 0. - Because the test used a mocked streaming surface rather than live endpoints, no HTTP endpoints or HTTP status messages were observed. <a href="https://app.greptile.com/trex/runs/13637854/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | core/anthropic/streaming/ledger.py | Allows final `message_delta` usage to override the ledger's initial estimated input token count. | | providers/deepseek/client.py | Reuses shared usage extraction for DeepSeek cache hit and miss token mapping. | | providers/transports/openai_chat/stream.py | Requests streamed usage and uses provider prompt and completion tokens when present. | | providers/transports/openai_chat/transport.py | Adds bounded retry behavior when upstream providers reject streamed usage metadata. | | providers/transports/openai_chat/usage.py | Adds shared helpers for usage requests, usage extraction, and usage rejection detection. | | tests/providers/test_openai_chat_usage.py | Adds coverage for streamed usage helpers, provider token usage, and retry fallback behavior. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant Client as Anthropic client participant Adapter as OpenAIChatStreamAdapter participant Transport as OpenAIChatTransport participant Provider as OpenAI-compatible provider participant Ledger as AnthropicStreamLedger Client->>Adapter: stream_response(request, input_tokens) Adapter->>Adapter: build body + request_stream_usage() Adapter->>Transport: _create_stream(body with include_usage) Transport->>Provider: "chat.completions.create(stream=True)" alt provider rejects include_usage Provider-->>Transport: 400/422 usage option error Transport->>Transport: clone_without_stream_usage() Transport->>Provider: retry without include_usage end Provider-->>Adapter: streaming chunks + optional final usage Adapter->>Adapter: usage_int(prompt_tokens/completion_tokens) Adapter->>Ledger: message_delta(input_tokens, output_tokens, provider usage fields) Ledger-->>Client: Anthropic SSE usage ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant Client as Anthropic client participant Adapter as OpenAIChatStreamAdapter participant Transport as OpenAIChatTransport participant Provider as OpenAI-compatible provider participant Ledger as AnthropicStreamLedger Client->>Adapter: stream_response(request, input_tokens) Adapter->>Adapter: build body + request_stream_usage() Adapter->>Transport: _create_stream(body with include_usage) Transport->>Provider: "chat.completions.create(stream=True)" alt provider rejects include_usage Provider-->>Transport: 400/422 usage option error Transport->>Transport: clone_without_stream_usage() Transport->>Provider: retry without include_usage end Provider-->>Adapter: streaming chunks + optional final usage Adapter->>Adapter: usage_int(prompt_tokens/completion_tokens) Adapter->>Ledger: message_delta(input_tokens, output_tokens, provider usage fields) Ledger-->>Client: Anthropic SSE usage ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Request streamed usage for OpenAI-chat p..."](https://github.com/alishahryar1/free-claude-code/commit/0473d8fe4bd042b5e247fc1dfa479afed9b60529) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=42593974)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
ccf46b88cf | Retry pre-stream provider transport failures (#1003) | ||
|
Ali Khokhar
|
bd85deb736 |
Fix OpenAI chat reasoning and tool history replay (#1002)
## Problem OpenAI-chat providers lost explicit empty reasoning state and could replay invalid tool-call history when unrelated messages appeared before matching tool results. ## Changes | Before | After | | --- | --- | | Empty `reasoning_content` and empty thinking blocks were treated as absent. | Empty reasoning is preserved as explicit replay state. | | OpenAI-chat conversion only deferred post-tool assistant text. | OpenAI-chat conversion buffers later transcript messages until required tool results are emitted. | | Responses prior tool calls and outputs were emitted one item per message. | Responses prior tool calls and outputs are grouped into valid Anthropic tool-use/result messages. | | Empty streamed `reasoning_content` produced no thinking block. | Empty streamed `reasoning_content` starts thinking state without visible delta text. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR fixes OpenAI chat reasoning replay and tool-history ordering. The main changes are: - Preserves explicit empty `reasoning_content` and empty thinking blocks. - Reworks OpenAI chat conversion around a ledger that waits for required tool results before replaying buffered transcript messages. - Groups prior Responses tool calls and outputs into valid Anthropic tool-use and tool-result messages. - Starts streamed thinking state when empty `reasoning_content` is received. - Adds focused tests for nested tool turns, out-of-order results, multi-tool replay, and empty reasoning. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with low risk. No blocking issues were found in the changed conversion paths. The updated ledger covers the prior invalid replay cases and the tests include nested, out-of-order, multi-tool, and empty reasoning scenarios. The required patch version and lockfile updates are present. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Ran the focused OpenAI conversion regression suite with Pytest, capturing the command, working directory, pass count, exit code, and elapsed time. - Encountered an external timeout during the initial Pytest run at 98% progress, then re-ran the same focused suite to completion for definitive proof. - Validated code quality with Ruff by executing the lint command and obtaining a successful output. <a href="https://app.greptile.com/trex/runs/13503694/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | core/anthropic/conversion.py | Replaces single pending-tool state with a ledger that buffers transcript segments until required OpenAI chat tool results can be emitted in valid order. | | core/openai_responses/input.py | Groups consecutive prior Responses tool calls/results into Anthropic tool-use/result turns and preserves explicit empty reasoning. | | core/openai_responses/reasoning.py | Updates reasoning extraction and combination helpers so empty strings remain explicit replay state without adding spurious separators. | | providers/deepseek/compat.py | Treats empty top-level or block-level thinking as replayable when detecting DeepSeek tool-history compatibility. | | providers/transports/openai_chat/stream.py | Starts an Anthropic thinking block for empty streamed `reasoning_content` while only emitting deltas for non-empty text. | | tests/providers/test_converter.py | Adds OpenAI chat conversion coverage for buffered tool history, nested pending tool turns, and explicit empty reasoning. | | tests/core/openai_responses/test_conversion.py | Adds Responses conversion tests for grouped prior tool calls/results and empty reasoning attachment. | | pyproject.toml | Bumps the package patch version for the production conversion fixes. | | uv.lock | Keeps the lockfile package version in sync with `pyproject.toml`. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant A as Anthropic transcript participant L as OpenAI chat ledger participant O as OpenAI chat history A->>L: Assistant tool_use segment L->>O: Emit assistant tool_calls A->>L: Later plain user/assistant messages L-->>L: Buffer until required tool_result ids arrive A->>L: User tool_result blocks L->>O: Emit matching role: tool results in tool_call order L->>O: Emit deferred assistant post-tool content L->>O: Drain buffered plain transcript messages ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant A as Anthropic transcript participant L as OpenAI chat ledger participant O as OpenAI chat history A->>L: Assistant tool_use segment L->>O: Emit assistant tool_calls A->>L: Later plain user/assistant messages L-->>L: Buffer until required tool_result ids arrive A->>L: User tool_result blocks L->>O: Emit matching role: tool results in tool_call order L->>O: Emit deferred assistant post-tool content L->>O: Drain buffered plain transcript messages ``` </a> </details> <sub>Reviews (3): Last reviewed commit: ["Refactor OpenAI chat tool history replay"](https://github.com/alishahryar1/free-claude-code/commit/ae1635d2ce3a232ba7f4c0b9787f7b604b625544) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=42274270)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
950aba393d |
Disable Hugging Face reasoning replay
Disable Hugging Face prior reasoning replay for Chat Completions while preserving streamed reasoning output. |
||
|
Ali Khokhar
|
62c0480eed | Add Mistral reasoning fallback | ||
|
Ali Khokhar
|
47ddedcada |
Fix NIM chat template downgrade (#997)
## Problem NVIDIA NIM Mistral-tokenizer models can reject chat-template controls with HTTP 400. FCC only removed `chat_template`, so requests that only had `chat_template_kwargs` still failed. ## Changes | Before | After | | --- | --- | | NIM retried chat-template errors by stripping only `extra_body.chat_template`. | NIM retries chat-template errors by stripping `extra_body.chat_template` and `extra_body.chat_template_kwargs`. | | Mistral-tokenizer models could fail after rejecting thinking chat-template kwargs. | Mistral-tokenizer models retry once without NIM chat-template controls. | | Reasoning-budget downgrades preserved thinking flags independently. | Reasoning-budget downgrades still preserve thinking flags independently. | | Package metadata stayed at `3.4.2`. | Package metadata is bumped to `3.4.3`. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR fixes the NVIDIA NIM chat-template retry downgrade. The main changes are: - Removes both `extra_body.chat_template` and `extra_body.chat_template_kwargs` before retrying chat-template 400 errors. - Adds tests for the full chat-template path and the kwargs-only regression case. - Adds helper coverage for unchanged request bodies. - Bumps package metadata from `3.4.2` to `3.4.3` and keeps `uv.lock` aligned. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with minimal risk. The production change is narrow and covered by targeted tests for both chat-template stripping paths. Version metadata and lockfile updates follow the repository guidance. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The initial focused chat-template test run was executed and showed 6 tests passed, while the outer shell wrapper returned nonzero due to a PIPESTATUS\[0\] substitution issue under /bin/sh after pytest completed. - A corrected wrapper rerun of the same focused chat-template tests was performed, and 6 tests passed with exit code 0. - Focused provider retry tests for chat\_template or reasoning\_budget were executed, and 4 tests passed with exit code 0. <a href="https://app.greptile.com/trex/runs/13364237/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | providers/nvidia_nim/retry.py | Extends NIM chat-template retry downgrades to remove both `chat_template` and `chat_template_kwargs` while preserving unchanged-body behavior. | | tests/providers/test_nvidia_nim.py | Updates streaming retry coverage to assert the second request strips all chat-template controls, including the kwargs-only regression case. | | tests/providers/test_nvidia_nim_request.py | Adds request-body helper coverage for stripping `chat_template_kwargs` and returning `None` when no chat-template controls are present. | | pyproject.toml | Bumps package version from `3.4.2` to `3.4.3` for the production bug fix. | | uv.lock | Keeps the editable package version in the lockfile aligned with `pyproject.toml`. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant Client as Claude request participant Provider as NvidiaNimProvider participant NIM as NVIDIA NIM API Client->>Provider: stream_response(request) Provider->>NIM: create(extra_body with chat_template controls) NIM-->>Provider: HTTP 400 chat_template rejection Provider->>Provider: clone_body_without_chat_template() Provider->>Provider: remove chat_template and chat_template_kwargs Provider->>NIM: retry create(extra_body without chat-template controls) NIM-->>Provider: stream chunks Provider-->>Client: SSE response events ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant Client as Claude request participant Provider as NvidiaNimProvider participant NIM as NVIDIA NIM API Client->>Provider: stream_response(request) Provider->>NIM: create(extra_body with chat_template controls) NIM-->>Provider: HTTP 400 chat_template rejection Provider->>Provider: clone_body_without_chat_template() Provider->>Provider: remove chat_template and chat_template_kwargs Provider->>NIM: retry create(extra_body without chat-template controls) NIM-->>Provider: stream chunks Provider-->>Client: SSE response events ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Fix NIM chat template downgrade"](https://github.com/alishahryar1/free-claude-code/commit/fb2d152372356e22441bd52aca99b1dca4792b61) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=42001002)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
15cab79a43 | Update GLM 5.2 model references | ||
|
newmemories360
|
770d56708a | Add SambaNova Cloud provider (#990) | ||
|
newmemories360
|
418f4963e5 | Fix HTTP 400 when max_(completion_)tokens exceeds a model's cap (#955) (#991) | ||
|
Ali Khokhar
|
0b86dd4ef8 |
Add GitHub Models provider (#989)
## Problem FCC does not expose GitHub Models, so users with GitHub Models access cannot route Claude, Codex, or messaging prompts through GitHub's hosted model catalog. ## Changes | Before | After | | --- | --- | | Provider catalog did not include GitHub Models. | Provider catalog includes `github_models` with token, proxy, admin, smoke, and model picker wiring. | | Requests could not target GitHub Models inference. | `providers/github_models` routes OpenAI-chat requests to `https://models.github.ai/inference`. | | Model discovery assumed provider `/models` compatibility. | GitHub Models discovery uses the catalog API and advertises stream/tool-capable models. | | OpenAI-chat transport could not set provider default headers. | OpenAI-chat transport accepts provider-owned default headers. | | Docs and templates omitted GitHub Models setup. | README, `.env.example`, and architecture docs document GitHub Models setup and ownership. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds GitHub Models as a new provider. The main changes are: - New `github_models` provider runtime, catalog, settings, and admin wiring. - OpenAI-chat transport support for provider-owned default headers. - GitHub Models catalog discovery filtered to streaming and tool-capable models. - Smoke configuration, environment template, docs, and tests for the new provider. - Package version and lockfile updates for the new feature. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with low risk. The provider is wired through runtime creation, catalog metadata, settings, admin fields, smoke config, docs, version metadata, and focused tests. No blocking correctness or security issues were found in the changed paths. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Before-change focused pytest run against HEAD^ showed no GitHub Models provider tests were collected. - After-change focused pytest run showed all 37 provider/runtime tests passed. - After-change harness output captured structured evidence for catalog discovery and OpenAI-chat request routing, and the harness exited successfully. - A temporary harness Python script was generated to capture the mocked request/response evidence. <a href="https://app.greptile.com/trex/runs/13317138/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | providers/github_models/client.py | Implements GitHub Models OpenAI-chat transport wiring, default GitHub headers, and catalog-based stream/tool-capable model discovery. | | providers/transports/openai_chat/transport.py | Allows OpenAI-chat providers to pass default headers into the shared AsyncOpenAI client. | | config/provider_catalog.py | Registers GitHub Models provider metadata, default inference base URL, credential, proxy, and capabilities. | | config/settings.py | Adds settings bindings for `GITHUB_MODELS_TOKEN` and `GITHUB_MODELS_PROXY`. | | api/admin_config/provider_manifest.py | Adds GitHub Models token labeling and description for generated admin provider fields. | | smoke/lib/config.py | Adds GitHub Models smoke defaults and credential detection. | | tests/providers/test_github_models.py | Adds focused tests for GitHub Models initialization, request conversion, catalog filtering, streaming, tool calls, reasoning, and cleanup. | | tests/providers/test_provider_runtime.py | Covers GitHub Models descriptor, provider config construction, and runtime instantiation. | | README.md | Adds GitHub Models setup documentation and updates provider counts/numbering. | | pyproject.toml | Bumps the package version to `3.2.0` for the new provider feature. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant User as Claude/Codex client participant FCC as FCC proxy/router participant Factory as Provider runtime factory participant GH as GitHubModelsProvider participant OpenAI as Shared OpenAI-chat transport participant API as models.github.ai User->>FCC: Request with model `github_models/...` FCC->>Factory: create_provider(`github_models`, settings) Factory->>GH: ProviderConfig(token, base_url, proxy) GH->>OpenAI: Initialize with GitHub default headers FCC->>GH: stream_response(MessagesRequest) GH->>OpenAI: build OpenAI chat body OpenAI->>API: "POST /inference/chat/completions (stream=true)" API-->>OpenAI: OpenAI-compatible stream chunks OpenAI-->>FCC: Anthropic SSE events FCC-->>User: Streamed Anthropic response FCC->>GH: list_model_infos() GH->>API: GET /catalog/models API-->>GH: Catalog entries with capabilities GH-->>FCC: stream/tool-capable model ids ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant User as Claude/Codex client participant FCC as FCC proxy/router participant Factory as Provider runtime factory participant GH as GitHubModelsProvider participant OpenAI as Shared OpenAI-chat transport participant API as models.github.ai User->>FCC: Request with model `github_models/...` FCC->>Factory: create_provider(`github_models`, settings) Factory->>GH: ProviderConfig(token, base_url, proxy) GH->>OpenAI: Initialize with GitHub default headers FCC->>GH: stream_response(MessagesRequest) GH->>OpenAI: build OpenAI chat body OpenAI->>API: "POST /inference/chat/completions (stream=true)" API-->>OpenAI: OpenAI-compatible stream chunks OpenAI-->>FCC: Anthropic SSE events FCC-->>User: Streamed Anthropic response FCC->>GH: list_model_infos() GH->>API: GET /catalog/models API-->>GH: Catalog entries with capabilities GH-->>FCC: stream/tool-capable model ids ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Add GitHub Models provider"](https://github.com/alishahryar1/free-claude-code/commit/736d3f9213f6a8d243d4001c5135b7fac402f143) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=41901295)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
9a17d1ed0a |
Add Cohere provider (#986)
## Problem FCC does not expose Cohere's OpenAI-compatible chat models, so users with Cohere keys cannot route Claude, Codex, or messaging prompts through Cohere. ## Changes | Before | After | | --- | --- | | Provider catalog did not include Cohere. | Provider catalog includes Cohere with `COHERE_API_KEY`, `COHERE_PROXY`, admin status, and smoke model wiring. | | Requests could not target Cohere's compatibility API. | `providers/cohere` routes OpenAI-chat requests to Cohere's compatibility API with Cohere-specific request policy. | | Docs and templates omitted Cohere setup. | README, `.env.example`, and architecture docs document Cohere setup and ownership. | | Cohere behavior had no regression coverage. | Provider, runtime, admin, config, smoke, and catalog tests cover Cohere integration. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Cohere as a new OpenAI-compatible chat provider. The main changes are: - Cohere provider metadata in the catalog, settings, Admin UI manifest, and runtime factory. - A new `CohereProvider` using the shared OpenAI chat transport with Cohere-specific request shaping. - Cohere API key, proxy, smoke model, README, architecture, and environment template updates. - Tests for Admin config, settings, provider catalog order, smoke config, runtime creation, and Cohere request/stream behavior. - Version and lockfile updates for the new provider feature. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with minimal risk. No functional, security, or contract issues were identified. Cohere is consistently wired through settings, catalog metadata, factory creation, Admin config, smoke defaults, docs, versioning, and targeted tests. The implemented Cohere `reasoning_effort` values match the Compatibility API behavior checked during review. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Validated the provider runtime handling of Cohere requests, including the request body policy, streaming parsing, and default base URL and API key behavior. - Verified that the runtime descriptor wiring and provider config proxy/key behavior pass in the general contract validation. - Confirmed the admin/config smoke contract artifact shows the Cohere environment settings, admin config masking, feature/provider catalog contracts, and smoke configuration passing. - Compared the initial -01-before.log and the clean -02-after.log captures to confirm the same scoped commands are present and both exit with code 0. <a href="https://app.greptile.com/trex/runs/13309850/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | providers/cohere/client.py | Implements Cohere request shaping over shared OpenAI chat transport, including allowed extra body and reasoning mapping; no issues found. | | config/provider_catalog.py | Registers Cohere with credential, proxy, base URL, transport, and capability metadata; no issues found. | | providers/runtime/factory.py | Wires Cohere into runtime provider factory dispatch; no issues found. | | config/settings.py | Adds Cohere API key and proxy settings aliases; no issues found. | | api/admin_config/provider_manifest.py | Adds Cohere API key labeling/description through catalog-derived Admin fields; no issues found. | | smoke/lib/config.py | Adds Cohere default smoke model and credential detection; no issues found. | | tests/providers/test_cohere.py | Adds request-policy and streaming adapter tests for the Cohere provider; no issues found. | | tests/providers/test_provider_runtime.py | Adds Cohere descriptor, config build, and factory instantiation coverage; no issues found. | | README.md | Adds Cohere setup instructions and updates provider counts/order; no issues found. | | pyproject.toml | Bumps package version for the new provider feature; no issues found. | | uv.lock | Updates the lockfile package version to match `pyproject.toml`; no issues found. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant User as User/Admin config participant Catalog as Provider catalog/settings participant Factory as Runtime factory participant Cohere as CohereProvider participant Transport as OpenAI chat transport participant API as Cohere Compatibility API User->>Catalog: "Configure MODEL=cohere/... and COHERE_API_KEY" Catalog->>Factory: Build ProviderConfig for provider_id cohere Factory->>Cohere: Instantiate CohereProvider(config) Cohere->>Transport: Build chat body with Cohere policy Transport->>API: "POST /chat/completions stream=true" API-->>Transport: Streaming OpenAI-compatible chunks Transport-->>User: Anthropic SSE events ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant User as User/Admin config participant Catalog as Provider catalog/settings participant Factory as Runtime factory participant Cohere as CohereProvider participant Transport as OpenAI chat transport participant API as Cohere Compatibility API User->>Catalog: "Configure MODEL=cohere/... and COHERE_API_KEY" Catalog->>Factory: Build ProviderConfig for provider_id cohere Factory->>Cohere: Instantiate CohereProvider(config) Cohere->>Transport: Build chat body with Cohere policy Transport->>API: "POST /chat/completions stream=true" API-->>Transport: Streaming OpenAI-compatible chunks Transport-->>User: Anthropic SSE events ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Add Cohere provider"](https://github.com/alishahryar1/free-claude-code/commit/7956802968fc6ce63bb71fe6d7503d484df56b79) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=41887611)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
d4683bf3f6 |
Add Hugging Face inference provider (#985)
## Problem FCC did not expose Hugging Face Inference Providers as a selectable backend. Voice transcription also used the legacy `HF_TOKEN` setting instead of the canonical Hugging Face API key. ## Changes | Before | After | | --- | --- | | Hugging Face models could not be selected through provider-prefixed routing. | Hugging Face routes through a thin OpenAI-chat provider using `huggingface/<model>`. | | Provider credentials did not include `HUGGINGFACE_API_KEY`. | Admin config, settings, smoke config, and docs use `HUGGINGFACE_API_KEY`. | | `HF_TOKEN` remained a voice-only config key. | Owned dotenv files migrate `HF_TOKEN` to `HUGGINGFACE_API_KEY`, while explicit `FCC_ENV_FILE` users get a warning. | | Version metadata stayed on `2.6.0`. | Version metadata moves to `3.0.0` with a refreshed lockfile. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Hugging Face Inference Providers as a selectable backend. The main changes are: - Adds a `huggingface` provider using the shared OpenAI-compatible chat transport. - Wires `HUGGINGFACE_API_KEY` and `HUGGINGFACE_PROXY` through settings, Admin UI, provider catalog, runtime factory, and smoke config. - Migrates owned dotenv files from `HF_TOKEN` to `HUGGINGFACE_API_KEY` and warns for explicit `FCC_ENV_FILE` users. - Updates voice transcription plumbing to use the canonical Hugging Face key. - Updates docs, examples, version metadata, lockfile, and related tests. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with minimal risk. No blocking correctness or security issues were identified. The new provider reuses the existing OpenAI-chat transport pattern. Provider wiring, env migration, Admin UI, smoke config, voice plumbing, and tests are consistent. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - The Pytest suite for providers, runtime, env migrations, config, and contract tests ran and completed with exit code 0 and 198 tests passed. - The HuggingFace runtime validator script ran and completed successfully, printing provider\_class=HuggingFaceProvider, default\_base\_url=https://router.huggingface.co/v1, credential\_env=HUGGINGFACE\_API\_KEY, and admin\_field=HUGGINGFACE\_API\_KEY:\[REDACTED\]. - Logs from both runs were captured as artifacts to aid review. <a href="https://app.greptile.com/trex/runs/13308618/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | providers/huggingface/client.py | Implements Hugging Face via the shared OpenAI-chat transport with `extra_body` passthrough. | | config/provider_catalog.py | Registers Hugging Face metadata, default router URL, credential, proxy, and capabilities. | | providers/runtime/factory.py | Wires Hugging Face into runtime provider construction. | | config/env_migrations.py | Adds safe `HF_TOKEN` to `HUGGINGFACE_API_KEY` dotenv migration helpers for owned env files. | | config/settings.py | Adds Hugging Face API key/proxy settings and removes the legacy `hf_token` setting. | | api/admin_config/manifest.py | Removes the voice-only `HF_TOKEN` field and adds Hugging Face smoke model configuration. | | api/admin_config/provider_manifest.py | Adds Admin UI labeling and description for `HUGGINGFACE_API_KEY`. | | messaging/transcription.py | Renames local Whisper token handling to use the canonical Hugging Face API key. | | smoke/lib/config.py | Adds Hugging Face smoke-test default model and credential detection. | | tests/providers/test_huggingface.py | Adds provider tests for Hugging Face base URL, request body policy, streaming, and cleanup. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant User as Admin/User participant Settings as Settings + dotenv migration participant Catalog as Provider Catalog participant Runtime as Provider Runtime Factory participant HF as HuggingFaceProvider participant Router as router.huggingface.co/v1 User->>Settings: "Configure MODEL=huggingface/<model> and HUGGINGFACE_API_KEY" Settings->>Settings: Rename owned HF_TOKEN to HUGGINGFACE_API_KEY when present Settings->>Catalog: Resolve huggingface descriptor and credential/proxy attrs Catalog->>Runtime: Build ProviderConfig for huggingface Runtime->>HF: Create HuggingFaceProvider HF->>Router: Stream OpenAI-compatible chat completion Router-->>HF: Streaming chunks HF-->>User: Anthropic SSE response ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant User as Admin/User participant Settings as Settings + dotenv migration participant Catalog as Provider Catalog participant Runtime as Provider Runtime Factory participant HF as HuggingFaceProvider participant Router as router.huggingface.co/v1 User->>Settings: "Configure MODEL=huggingface/<model> and HUGGINGFACE_API_KEY" Settings->>Settings: Rename owned HF_TOKEN to HUGGINGFACE_API_KEY when present Settings->>Catalog: Resolve huggingface descriptor and credential/proxy attrs Catalog->>Runtime: Build ProviderConfig for huggingface Runtime->>HF: Create HuggingFaceProvider HF->>Router: Stream OpenAI-compatible chat completion Router-->>HF: Streaming chunks HF-->>User: Anthropic SSE response ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Add Hugging Face inference provider"](https://github.com/alishahryar1/free-claude-code/commit/7341d9a923986ac84d5e4fdf858128f913f3e5d3) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=41885587)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
020bbef64b |
Add Vercel AI Gateway provider (#984)
## Problem FCC did not expose Vercel AI Gateway as a provider, so users with `AI_GATEWAY_API_KEY` could not route Claude, Codex, or messaging workflows through Vercel's model gateway. ## Changes | Before | After | | --- | --- | | Provider metadata skipped Vercel AI Gateway. | Provider metadata includes `vercel` with `AI_GATEWAY_API_KEY`, `VERCEL_AI_GATEWAY_PROXY`, and OpenAI-chat capabilities. | | No Vercel provider package or factory existed. | `VercelProvider` uses the shared OpenAI-chat transport with `max_tokens` and preserved `extra_body`. | | Admin, docs, smoke config, and model parsing had no Vercel surface. | Admin, docs, smoke config, and model parsing include Vercel model refs such as `vercel/openai/gpt-5.5`. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR adds Vercel AI Gateway as a new OpenAI-compatible provider. The main changes are: - Provider catalog, settings, Admin UI metadata, and runtime factory wiring for `vercel`. - A thin `VercelProvider` adapter that reuses the shared OpenAI-chat streaming transport. - Vercel-specific docs, environment examples, proxy settings, and smoke-test defaults. - Config, contract, runtime, and provider tests covering the new provider path. - Package version and lockfile updates for the new feature. </details> <h3>Confidence Score: 5/5</h3> This PR is safe to merge with minimal risk. The new provider follows the existing catalog, settings, factory, and shared transport patterns. The change includes focused config, runtime, smoke, and provider tests. The package version and lockfile were updated with the production changes. No functional or security issues were identified in the changed paths. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - A focused pytest run for the Vercel provider completed successfully with 187 tests passed in 3.93 seconds and EXIT\_CODE: 0. - An offline probe script named vercel-provider-offline-probe.py was generated to exercise the real factory/provider/request construction code offline. - The offline probe log vercel-provider-offline-probe.log showed the expected provider setup and a successful exit, including catalog\_has\_vercel=True, factory\_has\_vercel=True, provider class VercelProvider, base URL, synthetic API key propagation, max\_tokens preserved, max\_completion\_tokens absent, extra\_body preserved, and EXIT\_CODE: 0. <a href="https://app.greptile.com/trex/runs/13307261/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | README.md | Adds Vercel AI Gateway setup guidance and renumbers provider documentation. | | api/admin_config/provider_manifest.py | Adds Admin UI field metadata for `AI_GATEWAY_API_KEY` via existing catalog-derived manifest flow. | | config/provider_catalog.py | Registers `vercel` as an OpenAI-chat provider with gateway credential, default base URL, proxy, and capabilities. | | config/settings.py | Adds settings fields for Vercel gateway API key and proxy aliases. | | providers/runtime/factory.py | Wires the new `vercel` provider id to `VercelProvider` in runtime factory registration. | | providers/vercel/client.py | Implements a thin Vercel adapter over shared OpenAI-chat transport with `max_tokens` and `extra_body` passthrough. | | pyproject.toml | Bumps the package version to `2.6.0` for the new provider feature. | | smoke/lib/config.py | Adds Vercel smoke defaults and credential detection for provider smoke selection. | | tests/providers/test_provider_runtime.py | Adds runtime config and factory instantiation coverage for the Vercel provider. | | tests/providers/test_vercel.py | Adds unit tests for Vercel base URL handling, request-body policy, streaming deltas, and cleanup. | | uv.lock | Synchronizes the lockfile package version with `pyproject.toml`. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant User as User/Admin config participant Settings as Settings/env participant Catalog as Provider catalog participant Factory as Runtime factory participant Vercel as VercelProvider participant Gateway as Vercel AI Gateway User->>Settings: "Set AI_GATEWAY_API_KEY and MODEL=vercel/..." Settings->>Catalog: Resolve vercel descriptor Catalog->>Factory: Build ProviderConfig with key/base/proxy Factory->>Vercel: Instantiate VercelProvider Vercel->>Gateway: Stream OpenAI Chat Completions Gateway-->>Vercel: OpenAI-compatible chunks Vercel-->>User: Anthropic SSE via shared transport ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant User as User/Admin config participant Settings as Settings/env participant Catalog as Provider catalog participant Factory as Runtime factory participant Vercel as VercelProvider participant Gateway as Vercel AI Gateway User->>Settings: "Set AI_GATEWAY_API_KEY and MODEL=vercel/..." Settings->>Catalog: Resolve vercel descriptor Catalog->>Factory: Build ProviderConfig with key/base/proxy Factory->>Vercel: Instantiate VercelProvider Vercel->>Gateway: Stream OpenAI Chat Completions Gateway-->>Vercel: OpenAI-compatible chunks Vercel-->>User: Anthropic SSE via shared transport ``` </a> </details> <sub>Reviews (1): Last reviewed commit: ["Add Vercel AI Gateway provider"](https://github.com/alishahryar1/free-claude-code/commit/862ae946b2f7029ce4d3b309178312c9afe2751b) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=41883209)</sub> <!-- /greptile_comment --> |
||
|
Ali Khokhar
|
85b601884d |
Remove legacy future annotation imports (#982)
## Problem Python 3.14 provides native lazy annotations, but the codebase still relied on legacy future annotation imports. Those imports also made type-only import cycles easier to hide instead of fixing ownership boundaries. ## Changes | Before | After | | --- | --- | | Python files used `from __future__ import annotations`. | Python files rely on Python 3.14 native lazy annotations. | | Some runtime modules used `TYPE_CHECKING` or local imports for required dependencies. | Runtime modules use top-level owner-module imports with explicit boundaries. | | Local and GitHub guardrails only rejected type ignore suppressions. | Local and GitHub guardrails reject type ignore suppressions and legacy future annotation imports. | | Agent docs only documented the no-type-ignore rule. | Agent docs document the Python 3.14 annotation and import-boundary rules. | <!-- greptile_comment --> <details open><summary><h3>Greptile Summary</h3></summary> This PR moves the codebase to Python 3.14 native lazy annotations. The main changes are: - Removed legacy `from __future__ import annotations` imports across Python modules. - Promoted selected runtime dependencies from `TYPE_CHECKING` or local imports to explicit owner-module imports. - Added local, GitHub, and contract-test guardrails to reject legacy future annotation imports. - Updated agent docs with the annotation and import-boundary rules. - Bumped the package patch version for production-file changes. </details> <h3>Confidence Score: 5/5</h3> Safe to merge with low risk. The changes are mostly mechanical annotation cleanup with matching CI and contract-test guardrails. Reviewed import-boundary updates did not show a confirmed runtime cycle or dependency break. No files require special attention. <details><summary><h3><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="20" align="absmiddle"></a> T-Rex Logs</h3></summary> **What T-Rex did** - Performed an end-to-end validation of the guardrail contract suite: an environment check confirmed uv availability, a guardrail pytest run used CPython 3.14.0 with 5 passing contract tests, 3 focused CI-script tests passed, and the direct CI suppressions guardrail command (including the legacy future-annotations grep) also passed. <a href="https://app.greptile.com/trex/runs/13303335/artifacts"><picture><source media="(prefers-color-scheme: dark)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifactsDark.svg?v=4"><source media="(prefers-color-scheme: light)" srcset="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"><img alt="View all artifacts" src="https://greptile-static-assets.s3.amazonaws.com/badges/ViewAllArtifacts.svg?v=4"></picture></a> <sub><a href="https://www.greptile.com/trex"><img alt="T-Rex" src="https://greptile-static-assets.s3.amazonaws.com/trex/trex_green.svg" height="14" align="absmiddle"></a> Ran code and verified through T-Rex</sub> </details> <details open><summary><h3>Important Files Changed</h3></summary> | Filename | Overview | |----------|----------| | api/runtime.py | Moves messaging, CLI manager, session, limiter, and tree dependencies from local/type-checking imports to explicit top-level owner-module imports. | | messaging/platforms/telegram.py | Removes future annotations and promotes Telegram SDK type imports into the existing availability guard. | | messaging/platforms/telegram_inbound.py | Removes future annotations and imports Telegram SDK types at module scope for inbound normalization. | | tests/contracts/test_import_boundaries.py | Adds an AST contract that rejects legacy future annotation imports across Python files. | | scripts/ci.sh | Extends the local suppression check to reject legacy future annotation imports alongside type-ignore suppressions. | | scripts/ci.ps1 | Mirrors the local PowerShell CI suppression check for legacy future annotations. | | .github/workflows/tests.yml | Renames and broadens the GitHub guardrail job to reject both type suppressions and legacy future annotations. | | pyproject.toml | Bumps the patch version for production-file changes. | </details> <details open><summary><h3>Sequence Diagram</h3></summary> <a href="#gh-light-mode-only"> ```mermaid %%{init: {'theme': 'neutral'}}%% sequenceDiagram participant Dev as Developer/CI participant Guard as Suppression guard participant AST as Import-boundary contract test participant Py as Python modules Dev->>Guard: Run local/GitHub suppression check Guard->>Py: "Scan *.py for type ignores and future annotations" Guard-->>Dev: Fail if legacy annotation import remains Dev->>AST: Run pytest contract tests AST->>Py: Parse imports with ast AST-->>Dev: Assert no future annotations/import-boundary violations Py-->>Dev: Use Python 3.14 native lazy annotations ``` </a> <a href="#gh-dark-mode-only"> ```mermaid %%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%% sequenceDiagram participant Dev as Developer/CI participant Guard as Suppression guard participant AST as Import-boundary contract test participant Py as Python modules Dev->>Guard: Run local/GitHub suppression check Guard->>Py: "Scan *.py for type ignores and future annotations" Guard-->>Dev: Fail if legacy annotation import remains Dev->>AST: Run pytest contract tests AST->>Py: Parse imports with ast AST-->>Dev: Assert no future annotations/import-boundary violations Py-->>Dev: Use Python 3.14 native lazy annotations ``` </a> </details> <sub>Reviews (2): Last reviewed commit: ["Remove legacy future annotations import"](https://github.com/alishahryar1/free-claude-code/commit/6e6cda69da243bbdb92831207aecb3731ad469f8) | [Re-trigger Greptile](https://app.greptile.com/api/retrigger?id=41875785)</sub> <!-- /greptile_comment --> |