项目文件夹

文件
Seth Hobson be57c0b2e3 feat: multi-harness plugin marketplace (Codex, Cursor, OpenCode, Gemini) (#541)
* feat(adapters): multi-harness framework + harness_portability eval dimension

Turn this Claude Code plugin marketplace into a generic agentic-harness
marketplace. Adapters under tools/adapters/ emit harness-native artifacts
for OpenAI Codex CLI, Cursor, OpenCode, and Gemini CLI from a single
Markdown source. Source-of-truth stays under plugins/ — Claude Code is
unchanged.

Framework (tools/adapters/):
- base.py — PluginSource parser, HarnessAdapter ABC, write/mirror helpers
  (path-traversal guard, UTF-8-safe), inline-list + block-list + block-scalar
  YAML-ish parser, _utf8_safe_cut, _split_inline_list, _normalize_author
- capabilities.py — per-harness capability matrix, TOOL_NAME_MAPS,
  MODEL_ALIASES, resolve_model() with explicit warnings
- codex.py — emits .codex/{skills,agents}/ + AGENTS.md (≤150-line
  table-of-contents). Fence-aware body splitter, _utf8_safe_cut for
  multibyte safety, _yaml_scalar with reserved-word + special-char quoting.
  Skill/command name collision detection (and second-order __cmd fallback).
- cursor.py — emits .cursor-plugin/{plugin,marketplace}.json + curated
  .cursor/rules/*.mdc. _validate_mdc_frontmatter handles YAML block scalars
  (no false positives on colons in description body). _normalize_author
  handles dict, npm-style strings, and author lists.
- opencode.py — transpiles agents to .opencode/agents/<id>.md with
  mode:subagent + permission: deny-everything-else block (skill/task always
  allowed as base capabilities — Claude's implicit defaults).
- gemini.py — emits native skills/, agents/, and commands/ at extension
  root (April 2026 spec). Tool-allowlist remapped via TOOL_NAME_MAPS.

CLI + tooling:
- tools/generate.py — unified `make generate HARNESS=<x> [PLUGIN=<y>]`,
  with --clean (containment-guarded; case-insensitive on Darwin/Win32),
  --prune (orphan removal across all per-harness output trees), --strict
  (warnings fail), per-plugin error aggregation, refuses --clean --plugin
  (would silently wipe other plugins' artifacts).
- tools/validate_generated.py — structural validation across all four
  harness outputs. Codex 8KB cap → error. _extract_permission_block
  correctly handles nested permission keys (column-0 only).
- tools/doc_gardener.py — recurring drift detection per OpenAI harness-
  engineering principle. STALE_ARTIFACT (info), DEAD_LINK (error),
  MARKETPLACE_ORPHAN (error), SKILL_OVER_CODEX_CAP (warning), grouped
  output sorted by severity.

plugin-eval (extends existing framework):
- New harness_portability dimension (6% weight, rebalanced from existing
  static sub-scores). Surfaces non-portable patterns with concrete
  remediation hints: SKILL_OVER_CODEX_CAP, CLAUDE_TOOL_REFS,
  CLAUDE_TOOL_PROSE, AGENT_NAME_COLLISION, BARE_MODEL_ALIAS.
- _CAMEL_TOOL_PATTERN requires Claude-tool context (no false positives
  on Rust's `Task` etc.). _TOOL_PROSE_PATTERN case-sensitive on tool
  names, case-insensitive on the leading article.
- Findings do NOT also feed anti_pattern_penalty (no double-counting).

Documentation:
- Top-level guides: CODEX.md, CURSOR.md, OPENCODE.md (≤150 lines each,
  table-of-contents pattern per OpenAI harness-engineering post)
- docs/harnesses.md — capability matrix, graceful-degradation table,
  generated output paths
- docs/authoring.md — portable-content style guide (tools, models,
  collision rules, fence-respect)
- docs/round-trip-results.md — real-CLI verification recipes (OpenCode
  discovers 193 subagents, Gemini extensions validate passes, Codex
  TOMLs all parse)
- CONTRIBUTING.md — new file pointing at docs/authoring.md
- README.md — rewritten for multi-harness (145 lines, was 460)
- CLAUDE.md — trimmed to 60-line table-of-contents
- GEMINI.md — trimmed from 1500 to 500 tokens (3× over budget previously)

Tests: 181 passing (103 plugin-eval + 78 tools/tests). Real-CLI round-trip
verified for OpenCode, Gemini, and Codex (TOML parses).

Replaces tools/generate_gemini_commands.py with the unified CLI.

* refactor(skills): extract detail to references/details.md (~75 skills)

Apply Anthropic's canonical SKILL.md progressive-disclosure pattern across
the marketplace: SKILL.md body becomes a navigation tier (trigger phrasing
+ quick start), detailed templates and worked examples move to
references/details.md (loaded on demand by the agent).

Motivation: OpenAI Codex CLI hard-truncates skills at 8 KB. Before this
change, ~90 skills exceeded that cap and would silently break on Codex.
The progressive-disclosure pattern is also Anthropic's documented
recommendation for token efficiency — Claude Code reads references/ files
on demand when the body navigation says to.

What's extracted, by pattern:
- Pass 1 (## Templates section): 19 skills — full template libraries
  moved to references/details.md
- Pass 2 (## Implementation Patterns / ## Advanced Patterns): 13 skills
- Pass 3 (everything between nav-tier and wrap-tier headings): 53 skills
- Conservative re-extraction for 8 skills that got over-reduced — kept
  ~6-7 KB inline (most of the quick-start tier) plus references/ overflow

What stays inline (SKILL.md navigation tier):
- description: frontmatter (triggering — unchanged for all skills)
- ## When to Use This Skill / ## Core Concepts / ## Quick Start
- ## Best Practices / ## Troubleshooting / ## See Also wrap-ups
- A pointer note ("see references/details.md") so the agent knows where
  to look for detail

What goes to references/details.md (detail tier, on-demand load):
- ## Templates (full code template libraries)
- ## Implementation Patterns / ## Advanced Patterns (deep examples)
- Mid-skill walkthroughs that exceed the inline budget

Also in this commit:
- plugins/brand-landingpage description trimmed from 958→543 chars
  (preserves trigger phrasing, drops verbose example-quote list)

Net effect:
- SKILL_OVER_CODEX_CAP findings: 90 → 10 (88% reduction)
- All triggers unchanged — discovery behavior identical across harnesses
- 75 new references/details.md files with the extracted content
- Same depth of guidance, loaded progressively

Remaining 10 oversized skills are complex multi-section docs (e.g.
postgresql, code-review-excellence, evaluation-methodology) that need
per-skill manual judgment — flagged by `make garden` for future work.

* chore: bump all plugin versions (multi-harness release)

Patch-bump every local plugin (81) in both .claude-plugin/marketplace.json
entries and each plugins/<name>/.claude-plugin/plugin.json. Minor-bump the
top-level marketplace metadata.version (1.6.0 → 1.7.0) to signal the
multi-harness adapter framework addition.

The external git-subdir entry (qa-orchestra) is unaffected — its version
is governed by its upstream repo.

* fix(opencode): preserve explicit tools:[] + word-boundary subtask match

Addresses two Codex review findings on PR #541.

## P1 — `tools: []` silently upgraded to permissive (privilege escalation)

Before: `_build_permission_block` returned `{}` for any empty list, which
omits the `permission:` block entirely from the emitted agent. An author
who explicitly wrote `tools: []` to lock down an advisory-only agent got
an UNRESTRICTED agent in OpenCode. Affected agent in this tree:
`plugins/arm-cortex-microcontrollers/agents/arm-cortex-expert.md`.

Fix: `_build_permission_block` now takes a `has_tools_field` flag so the
caller can distinguish "tools: key missing" (Claude default permissive)
from "tools: []" (explicit lock-down). The lock-down case emits a
deny-everything block that allows ONLY the base capabilities (skill, task)
that Claude Code always grants implicitly. Verified against the real
arm-cortex-expert agent — now emits read/edit/write/bash/grep/glob/list:
deny, task/skill: allow.

## P2 — `"agent" in cmd.body.lower()` false-positives on substrings

Before: a command body containing `PerformanceReviewAgent` (class name
in a code snippet) or `useragent` triggered `subtask: true`, changing
runtime behavior based on incidental text.

Fix: switch to a compiled word-boundary regex `\b(agent|subagent)s?\b`
(case-insensitive). Tests confirm the substring `PerformanceReviewAgent`
no longer fires, while a real "spawn a subagent" sentence still does.

## Tests

3 new regression tests in tools/tests/test_adapters.py:
- `test_explicit_empty_tools_yields_locked_permission_block` (P1)
- `test_missing_tools_field_yields_no_permission_block` (P1 boundary)
- `test_subtask_inference_word_boundary` (P2)

184 total tests pass (was 181). OpenCode round-trip still discovers all
193 subagents; arm-cortex-expert agent is now properly locked down.

* test: behavioral verification + CI gates for multi-harness pipeline

Adds three layers of automated verification that pure-Python parser tests
miss, plus the CI jobs that turn them into hard gates. Catches the kinds
of issues that previously only surfaced when a real user installed the
marketplace and tried to use it.

## test_real_world.py — real-source structural tests

Runs against the actual `plugins/` tree (not synthetic fixtures). Catches
issues that only appear on real content:

- every marketplace entry resolves to a plugins/<name>/ dir
- every local plugin dir appears in marketplace.json
- marketplace.json version == per-plugin plugin.json version (catches drift)
- every plugin loads via load_plugin() without error
- no plugin name contains `__` (adapter namespace separator)
- every agent has name + description; every skill has a trigger phrase
  (same regex plugin_eval's MISSING_TRIGGER check uses)
- no agent name collides with Codex built-ins
- every refactored skill (with `references/details.md`) has:
  - meaningful detail content (>=500 B in details.md)
  - a pointer to references/ in the SKILL.md body
  - a navigation-tier heading preserved (When to Use, Overview, etc.)
  - body >= 600 B (not a stub)
- every plugin.json has name + version matching the dir

This test pass found and fixed three real defects before commit:
- ship-mate/skills/scan: description had no trigger phrase ("Use when…")
- reverse-engineering/skills/memory-forensics: nav-tier section lost
  during extraction
- reverse-engineering/skills/binary-analysis-patterns: same

All three are now fixed (preserved trigger phrasing, added When-to-Use
sections back to the skills my extraction over-trimmed).

## test_round_trip.py — generate→parse→verify

CI runs this AFTER `make generate-all`. Catches generation-time regressions:

- OpenCode/Codex/Gemini agent counts match source agent count (no skips)
- every Codex SKILL.md under 8 KB (the cap that would silently truncate)
- every Codex agent TOML has required fields + valid sandbox_mode
- every OpenCode agent has mode in {primary,subagent,all} and
  provider-prefixed model
- locked agents (source `tools: []`) emit proper deny-everything permission
  block with skill/task allow (regression guard for PR-541 P1)
- every Gemini @{path} injection resolves to a real source file
- every Gemini command TOML has prompt + {{args}} placeholder
- every context file (CLAUDE.md, AGENTS.md, GEMINI.md, etc.) within
  150-line cap
- Cursor marketplace + per-plugin manifests cover all local plugins
- .cursor/rules/*.mdc only use the 3 documented frontmatter keys

## test_cli_smoke.py — real-CLI subprocess tests

Invokes the actual harness binaries (OpenCode, Gemini, Codex, Claude Code)
against the generated artifacts. Catches CLI-level issues pure-Python
parsing can't see: schema-loader drift, plugin-discovery bugs, version
incompatibilities.

- `opencode agent list` — must succeed AND discover every source agent
  (currently 191 + 2 OpenCode built-ins)
- `gemini extensions validate <repo>` — must return success
- `codex doctor` — must report healthy install
- every Codex agent TOML must parse with stdlib `tomllib`
- `claude --version` — sanity check the Claude Code CLI loads
- marketplace.json must have owner + metadata.version for Claude Code's loader

Per-CLI tests skip gracefully when the binary isn't on PATH, so local
devs only exercise what they have installed. CI installs OpenCode +
Gemini and turns those skips into hard gates.

## Makefile + CI

- `make test` — full pytest suite (plugin-eval + tools/tests/)
- `make smoke-test` — generates if needed, then runs real-CLI smoke tests
- `.github/workflows/validate.yml` extended with:
  - `tools-tests` job — runs pytest tools/tests/
  - `multi-harness-generate` job — `make generate-all && make validate
    STRICT=1 && make garden`, uploads generated artifacts on every run
  - `cli-smoke-test` job — installs OpenCode + Gemini, runs test_cli_smoke.py

## Test counts

- Before: 184 tests
- After: 386 tests (parameterized real-source tests over all 82 plugins)
- All passing locally on OpenCode 1.15.7 + Gemini 0.42.0 + Codex 0.133.0
  + Claude Code 2.1.148
2026-05-22 08:18:21 -04:00

9.7 KiB

rust-async-patterns — detailed patterns and worked examples

Patterns

Pattern 1: Concurrent Task Execution

use tokio::task::JoinSet;
use anyhow::Result;

// Spawn multiple concurrent tasks
async fn fetch_all_concurrent(urls: Vec<String>) -> Result<Vec<String>> {
    let mut set = JoinSet::new();

    for url in urls {
        set.spawn(async move {
            fetch_data(&url).await
        });
    }

    let mut results = Vec::new();
    while let Some(res) = set.join_next().await {
        match res {
            Ok(Ok(data)) => results.push(data),
            Ok(Err(e)) => tracing::error!("Task failed: {}", e),
            Err(e) => tracing::error!("Join error: {}", e),
        }
    }

    Ok(results)
}

// With concurrency limit
use futures::stream::{self, StreamExt};

async fn fetch_with_limit(urls: Vec<String>, limit: usize) -> Vec<Result<String>> {
    stream::iter(urls)
        .map(|url| async move { fetch_data(&url).await })
        .buffer_unordered(limit) // Max concurrent tasks
        .collect()
        .await
}

// Select first to complete
use tokio::select;

async fn race_requests(url1: &str, url2: &str) -> Result<String> {
    select! {
        result = fetch_data(url1) => result,
        result = fetch_data(url2) => result,
    }
}

Pattern 2: Channels for Communication

use tokio::sync::{mpsc, broadcast, oneshot, watch};

// Multi-producer, single-consumer
async fn mpsc_example() {
    let (tx, mut rx) = mpsc::channel::<String>(100);

    // Spawn producer
    let tx2 = tx.clone();
    tokio::spawn(async move {
        tx2.send("Hello".to_string()).await.unwrap();
    });

    // Consume
    while let Some(msg) = rx.recv().await {
        println!("Got: {}", msg);
    }
}

// Broadcast: multi-producer, multi-consumer
async fn broadcast_example() {
    let (tx, _) = broadcast::channel::<String>(100);

    let mut rx1 = tx.subscribe();
    let mut rx2 = tx.subscribe();

    tx.send("Event".to_string()).unwrap();

    // Both receivers get the message
    let _ = rx1.recv().await;
    let _ = rx2.recv().await;
}

// Oneshot: single value, single use
async fn oneshot_example() -> String {
    let (tx, rx) = oneshot::channel::<String>();

    tokio::spawn(async move {
        tx.send("Result".to_string()).unwrap();
    });

    rx.await.unwrap()
}

// Watch: single producer, multi-consumer, latest value
async fn watch_example() {
    let (tx, mut rx) = watch::channel("initial".to_string());

    tokio::spawn(async move {
        loop {
            // Wait for changes
            rx.changed().await.unwrap();
            println!("New value: {}", *rx.borrow());
        }
    });

    tx.send("updated".to_string()).unwrap();
}

Pattern 3: Async Error Handling

use anyhow::{Context, Result, bail};
use thiserror::Error;

#[derive(Error, Debug)]
pub enum ServiceError {
    #[error("Network error: {0}")]
    Network(#[from] reqwest::Error),

    #[error("Database error: {0}")]
    Database(#[from] sqlx::Error),

    #[error("Not found: {0}")]
    NotFound(String),

    #[error("Timeout after {0:?}")]
    Timeout(std::time::Duration),
}

// Using anyhow for application errors
async fn process_request(id: &str) -> Result<Response> {
    let data = fetch_data(id)
        .await
        .context("Failed to fetch data")?;

    let parsed = parse_response(&data)
        .context("Failed to parse response")?;

    Ok(parsed)
}

// Using custom errors for library code
async fn get_user(id: &str) -> Result<User, ServiceError> {
    let result = db.query(id).await?;

    match result {
        Some(user) => Ok(user),
        None => Err(ServiceError::NotFound(id.to_string())),
    }
}

// Timeout wrapper
use tokio::time::timeout;

async fn with_timeout<T, F>(duration: Duration, future: F) -> Result<T, ServiceError>
where
    F: std::future::Future<Output = Result<T, ServiceError>>,
{
    timeout(duration, future)
        .await
        .map_err(|_| ServiceError::Timeout(duration))?
}

Pattern 4: Graceful Shutdown

use tokio::signal;
use tokio::sync::broadcast;
use tokio_util::sync::CancellationToken;

async fn run_server() -> Result<()> {
    // Method 1: CancellationToken
    let token = CancellationToken::new();
    let token_clone = token.clone();

    // Spawn task that respects cancellation
    tokio::spawn(async move {
        loop {
            tokio::select! {
                _ = token_clone.cancelled() => {
                    tracing::info!("Task shutting down");
                    break;
                }
                _ = do_work() => {}
            }
        }
    });

    // Wait for shutdown signal
    signal::ctrl_c().await?;
    tracing::info!("Shutdown signal received");

    // Cancel all tasks
    token.cancel();

    // Give tasks time to cleanup
    tokio::time::sleep(Duration::from_secs(5)).await;

    Ok(())
}

// Method 2: Broadcast channel for shutdown
async fn run_with_broadcast() -> Result<()> {
    let (shutdown_tx, _) = broadcast::channel::<()>(1);

    let mut rx = shutdown_tx.subscribe();
    tokio::spawn(async move {
        tokio::select! {
            _ = rx.recv() => {
                tracing::info!("Received shutdown");
            }
            _ = async { loop { do_work().await } } => {}
        }
    });

    signal::ctrl_c().await?;
    let _ = shutdown_tx.send(());

    Ok(())
}

Pattern 5: Async Traits

use async_trait::async_trait;

#[async_trait]
pub trait Repository {
    async fn get(&self, id: &str) -> Result<Entity>;
    async fn save(&self, entity: &Entity) -> Result<()>;
    async fn delete(&self, id: &str) -> Result<()>;
}

pub struct PostgresRepository {
    pool: sqlx::PgPool,
}

#[async_trait]
impl Repository for PostgresRepository {
    async fn get(&self, id: &str) -> Result<Entity> {
        sqlx::query_as!(Entity, "SELECT * FROM entities WHERE id = $1", id)
            .fetch_one(&self.pool)
            .await
            .map_err(Into::into)
    }

    async fn save(&self, entity: &Entity) -> Result<()> {
        sqlx::query!(
            "INSERT INTO entities (id, data) VALUES ($1, $2)
             ON CONFLICT (id) DO UPDATE SET data = $2",
            entity.id,
            entity.data
        )
        .execute(&self.pool)
        .await?;
        Ok(())
    }

    async fn delete(&self, id: &str) -> Result<()> {
        sqlx::query!("DELETE FROM entities WHERE id = $1", id)
            .execute(&self.pool)
            .await?;
        Ok(())
    }
}

// Trait object usage
async fn process(repo: &dyn Repository, id: &str) -> Result<()> {
    let entity = repo.get(id).await?;
    // Process...
    repo.save(&entity).await
}

Pattern 6: Streams and Async Iteration

use futures::stream::{self, Stream, StreamExt};
use async_stream::stream;

// Create stream from async iterator
fn numbers_stream() -> impl Stream<Item = i32> {
    stream! {
        for i in 0..10 {
            tokio::time::sleep(Duration::from_millis(100)).await;
            yield i;
        }
    }
}

// Process stream
async fn process_stream() {
    let stream = numbers_stream();

    // Map and filter
    let processed: Vec<_> = stream
        .filter(|n| futures::future::ready(*n % 2 == 0))
        .map(|n| n * 2)
        .collect()
        .await;

    println!("{:?}", processed);
}

// Chunked processing
async fn process_in_chunks() {
    let stream = numbers_stream();

    let mut chunks = stream.chunks(3);

    while let Some(chunk) = chunks.next().await {
        println!("Processing chunk: {:?}", chunk);
    }
}

// Merge multiple streams
async fn merge_streams() {
    let stream1 = numbers_stream();
    let stream2 = numbers_stream();

    let merged = stream::select(stream1, stream2);

    merged
        .for_each(|n| async move {
            println!("Got: {}", n);
        })
        .await;
}

Pattern 7: Resource Management

use std::sync::Arc;
use tokio::sync::{Mutex, RwLock, Semaphore};

// Shared state with RwLock (prefer for read-heavy)
struct Cache {
    data: RwLock<HashMap<String, String>>,
}

impl Cache {
    async fn get(&self, key: &str) -> Option<String> {
        self.data.read().await.get(key).cloned()
    }

    async fn set(&self, key: String, value: String) {
        self.data.write().await.insert(key, value);
    }
}

// Connection pool with semaphore
struct Pool {
    semaphore: Semaphore,
    connections: Mutex<Vec<Connection>>,
}

impl Pool {
    fn new(size: usize) -> Self {
        Self {
            semaphore: Semaphore::new(size),
            connections: Mutex::new((0..size).map(|_| Connection::new()).collect()),
        }
    }

    async fn acquire(&self) -> PooledConnection<'_> {
        let permit = self.semaphore.acquire().await.unwrap();
        let conn = self.connections.lock().await.pop().unwrap();
        PooledConnection { pool: self, conn: Some(conn), _permit: permit }
    }
}

struct PooledConnection<'a> {
    pool: &'a Pool,
    conn: Option<Connection>,
    _permit: tokio::sync::SemaphorePermit<'a>,
}

impl Drop for PooledConnection<'_> {
    fn drop(&mut self) {
        if let Some(conn) = self.conn.take() {
            let pool = self.pool;
            tokio::spawn(async move {
                pool.connections.lock().await.push(conn);
            });
        }
    }
}

Debugging Tips

// Enable tokio-console for runtime debugging
// Cargo.toml: tokio = { features = ["tracing"] }
// Run: RUSTFLAGS="--cfg tokio_unstable" cargo run
// Then: tokio-console

// Instrument async functions
use tracing::instrument;

#[instrument(skip(pool))]
async fn fetch_user(pool: &PgPool, id: &str) -> Result<User> {
    tracing::debug!("Fetching user");
    // ...
}

// Track task spawning
let span = tracing::info_span!("worker", id = %worker_id);
tokio::spawn(async move {
    // Enters span when polled
}.instrument(span));