main
2 次代码提交
| 作者 | SHA1 | 备注 | 提交日期 | |
|---|---|---|---|---|
 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
|
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 --> |