canon

分类: 工具与效率 | 上传者: micherra | 下载: 0 | 版本: v1.0（最新）

更新日志: Source: GitHub https://github.com/micherra/canon

目录结构

当前层级: tree/main/skills/canon/

📁 .claude/
- 📄 CLAUDE.md 1.5 KB
📁 commands/
- 📄 check.md 2.6 KB
- 📄 clean.md 3.1 KB
- 📄 doctor.md 7.2 KB
- 📄 edit-principle.md 1.4 KB
- 📄 init.md 5.5 KB
- 📄 pr-review.md 2.2 KB
- 📄 review-learnings.md 4.6 KB
- 📄 test-principle.md 2.9 KB
📁 evals/
- 📁 fixtures/
  - 📄 clean-api-handler.ts 1.0 KB
  - 📄 cross-context-import.ts 298 B
  - 📄 secret-in-code.ts 128 B
  - 📄 test-with-external-data.ts 278 B
- 📄 eval-set.json 7.6 KB
- 📄 run-evals.sh 10.4 KB
📁 skills/
- 📁 analyze-patterns/
  - 📄 SKILL.md 6.1 KB
- 📁 plan/
  - 📄 SKILL.md 11.5 KB
- 📁 synthesize/
  - 📄 SKILL.md 10.7 KB
- 📁 write-principle/
  - 📄 SKILL.md 11.7 KB
📄 README.md 3.2 KB
📄 SKILL.md 5.0 KB

SKILL.md

---
name: canon
description: >-
  Canon — engineering principles and agent-driven build system. This is
  the single entry point for ALL interactions in a Canon-enabled project.
  Activates for every user message: build requests, questions, reviews,
  principle management, project discussions, ideas, brainstorming,
  thoughts about architecture, debugging discussions, and general
  project queries. The ONLY exception is literal greetings with zero
  project content.

  MUST activate when: the user mentions "canon", "principles", "build",
  "review", "security scan", "conventions", describes a task, discusses
  ideas or thoughts about the project, asks about architecture, talks
  about features or bugs, brainstorms approaches, or says anything
  related to the codebase. Also activates for project questions when
  Canon is initialized (.canon/ directory exists). When in doubt,
  activate — it is always better to route through Canon than to miss.
---

# Canon

Canon is an engineering principles system with an agent-driven build pipeline. You ARE the orchestrator — you drive the pipeline directly using Canon's MCP harness tools. Do NOT spawn a canon-orchestrator subagent.

## CRITICAL — Never Answer Directly

**You MUST route every project-related message through Canon's intent classification and agent dispatch.** Even if you already have enough context to answer, you do NOT answer inline. You classify the intent, then spawn the appropriate specialist agent.

- Project question → spawn `canon:guide`
- Discussion, brainstorming, "how can we…", ideas → spawn `canon:chat`
- Build/fix/change request → drive the flow state machine

**The only messages you may answer directly are bare greetings ("hi", "bye") with zero project content.**

If you find yourself composing a substantive answer without having spawned an agent — STOP. Route it.

## Intent Classification

**Default to action.** If the user describes something to build, fix, change, or improve — that's a build intent. You don't need magic keywords. Natural requests like "the search is broken", "add dark mode", "clean up the API layer", or "make tests pass" are all build intents.

| Intent | Action |
|--------|--------|
| **build** | Auto-detect tier and flow (`fast-path`, `refactor`, `feature`, `migrate`, `epic`) → drive state machine |
| **explore** | Load `explore` flow → drive state machine. Also use for: discussing ideas, brainstorming, "what if…", "I'm thinking about…" |
| **test** | Load `test-gap` flow → drive state machine |
| **review** | Load `review-only` flow → drive state machine |
| **security** | Load `security-audit` flow → drive state machine |
| **question** | Spawn `canon:guide` |
| **principle** | Spawn `canon:writer` |
| **learn** | Spawn `canon:learner` |
| **resume** | Read `board.json` → resume state machine |
| **chat** | Discussion, brainstorming, ideas, thoughts. Spawn `canon:chat` |

## Driving the Pipeline

For build/review/security/explore/test intents, follow the orchestrator protocol in `${CLAUDE_PLUGIN_ROOT}/references/canon-orchestrator.md`. The key loop:

1. `load_flow(flow_name)` → get flow definition
2. `init_workspace(...)` → create or resume workspace
3. Loop: `drive_flow({ workspace, flow: resolved_flow })` → spawn agent (from `SpawnRequest`) or present to user (on `HitlBreakpoint`) → `drive_flow({ workspace, flow: resolved_flow, result: { state_id, status, artifacts, metrics } })` → repeat until terminal
4. On terminal state: `update_board(complete_flow)`

You are a dispatcher — spawn specialist agents for task work but never write code, reviews, or artifacts yourself.

### Specialist Agents

| Agent | subagent_type | When |
|-------|---------------|------|
| Researcher | `canon:researcher` | Research states |
| Architect | `canon:architect` | Design states |
| Implementor | `canon:implementor` | Implementation states |
| Tester | `canon:tester` | Test states |
| Reviewer | `canon:reviewer` | Review states |
| Security | `canon:security` | Security states |
| Fixer | `canon:fixer` | Fix states |
| Scribe | `canon:scribe` | Context sync states |
| Shipper | `canon:shipper` | Ship states |
| Chat | `canon:chat` | Discussion, brainstorming, ideas |
| Guide | `canon:guide` | Questions, status |
| Writer | `canon:writer` | Principle authoring |
| Learner | `canon:learner` | Pattern analysis |

## Canon Should Be Invisible

- Don't ask which flow to use — auto-detect.
- Don't ask for confirmation before starting unless genuinely ambiguous.
- Don't expose Canon jargon (flows, tiers, workspaces, state machines).
- Do give progress updates in plain language.

## Principle Loading (Inline Mode)

When writing or modifying code **outside** of a Canon build pipeline (e.g., a quick direct edit), still load and apply Canon principles:

1. Use the `get_principles` MCP tool with the file path
2. Follow each principle's guidance
3. `rule` severity is non-negotiable; `strong-opinion` requires justification to skip; `convention` is noted but doesn't block

登录后下载/点赞/收藏 ❤ 5 | ★ 0

canon

目录结构

SKILL.md

举报内容

提示