[Documentation] Use when generating, smart-merging or refactoring the root AI context (CLAUDE.md, AGENTS.md) and its Claude/Codex projections. Config file only: project-config.
日本語の概要は準備中です。原文の説明を表示しています。
[Documentation] Use when a workflow step or the user asks for documentation sync after code/spec/test changes (--mode=update), or explicit reference-doc initialization and reconciliation (--mode=init).
日本語の概要は準備中です。原文の説明を表示しています。
インストール方法を見るインストールする前に、エージェントに与えられる指示の中身を確認できます。
[BLOCKING] Mode routing — detect FIRST. Explicit
--mode=updateor--mode=initselects that mode./docs-manager --mode=updateand/docs-manager --mode=initare the former/docs-updateand/docs-init: those slash commands no longer exist, and each mode works called directly with no workflow. No mode flag and no natural-language request matching theupdateIntent column: show the Mode Dispatch table and stop — ask nothing and run nothing (initis heavy and runs only on an explicit--mode=init). Read the mode file in full before anything else.
Goal: Keep documentation true to code: update impacted context, config, specs/tests and derived/demo outputs; explicitly initialize or reconcile selected reference docs.
Summary: Update impacted docs or explicitly initialize reference docs: detect mode → read its contract → resolve scope/owners → run gates → review and report evidence.
Workflow: Detect mode → read its full reference → execute its contract → verify and report.
Key Rules:
update Intent column selects update; init is selected only by an explicit --mode=init. No mode flag and no such request: show the Mode Dispatch table and stop.--mode= (two dashes) selects the skill mode. The update caller flag mode=update (no dashes, see that mode's Scope and inputs) only overrides /spec mode detection and never selects a skill mode.docs-manager sub-agent (subagent_type="docs-manager") is an agent, not this skill; it drives /docs-manager --mode=update.file:line, confidence >80% to act) and task tracking live as each step starts and completes.| Mode | Purpose | Intent (natural-language triggers) | Read in full FIRST |
|---|---|---|---|
--mode=update [modules=… changed_files=… phases=… skip_phases=… tc_mode=… freshness={impact|full|off} base=…] | Impact-scoped context/config checks → spec/test owners → derived/demo outputs → report. Formerly /docs-update | "update docs", "docs impacted by my changes", "sync docs after code change", "doc sync", "documentation update" | references/mode-update.md |
--mode=init | Initialize or reconcile selected, applicable reference docs through scan. Formerly /docs-init | (explicit --mode=init only) | references/mode-init.md |
--mode=update, read references/mode-update.md in full FIRST; it owns scope, owners, freshness, intent/coverage, caller flags and the report (tmp/reports/docs-update-{YYMMDD}-{HHMM}.md). Workflow invocation and standalone both run the full contract.--mode=init, read references/mode-init.md in full FIRST; it owns optional config, document selection and applicability-gated scans.scan (/scan --target=<key>); refreshing every selected reference doc at once stays in scan-all. --mode=update escalates to them and --mode=init delegates to scan; neither is a mode of this skill.Protocol guides — Apply each protocol only when its read-when condition holds. If a protocol's text is not in your context, read its file below before you act on it.
output-quality-principles — Readable guidance and verified contents for long skill/doc files; writing, updating or auditing a skill or reference document, or another Markdown artifact over 200 lines → .claude/skills/shared/protocols/output-quality-principles.mdIMPORTANT MUST ATTENTION lead with useful guidance and readable priorities; preserve action-changing conditions/numbers and required structures. When writing, updating or auditing Markdown skill/doc content, verify top, heading-linked contents over 200 lines (stricter owner thresholds remain); refresh after edits and report gaps in read-only audits. Remove report bulk; judge semantic value rather than word counts.
<!-- /SYNC:output-quality-principles:reminder -->IMPORTANT MUST ATTENTION Goal: Keep project documentation true to the code through the selected update or init contract.
IMPORTANT MUST ATTENTION Main steps: detect mode → read its full reference → resolve scope and ownership → run quality gates → review and report evidence.
--mode=<x> reads references/mode-<x>.md in full FIRST; no mode flag and no update-intent request shows the mode table and stops — never guess a mode, never start init without an explicit --mode=initfile:line evidence for every claim (confidence >80% to act)[TASK-PLANNING] Track the selected contract before acting, with a final consistency review.
| Evasion | Required action |
|---|---|
| "Direct call, skip the reference" | Standalone calls retain the full selected contract. |
まだレビューはありません。使ってみた感想をお寄せください。
ログインなしで書き込めます。使ってみた感想、質問、動かなかった環境などを気軽にどうぞ。
まだコメントはありません。最初のコメントを書いてみませんか。
概要と使いどころ
[Documentation] Use when generating, smart-merging or refactoring the root AI context (CLAUDE.md, AGENTS.md) and its Claude/Codex projections. Config file only: project-config.
日本語の概要は準備中です。原文の説明を表示しています。
[Code Quality] Use when a workflow step or the user asks for AI code or plan review: LLM calls, prompts, agents, RAG, evals and guardrails. --mode={code|plan}; --report-only is read-only.
日本語の概要は準備中です。原文の説明を表示しています。
[Architecture] Use when a workflow step or the user asks for architecture design, compliance review, scalability grading or whole-project audit via --mode={design|review|scalability|full}.
日本語の概要は準備中です。原文の説明を表示しています。
[Content] Use when a workflow step or the user asks for product/feature ideation with ranked opportunities and validation tests. --mode={roadmap|scope}.
日本語の概要は準備中です。原文の説明を表示しています。
[Content] Use when a workflow step or the user asks for business viability: Business Model Canvas, financials, risks and go-to-market recommendation.
日本語の概要は準備中です。原文の説明を表示しています。
[Code Quality] Use when a workflow step or the user asks for a focused code-change review: staged, unstaged, branch diff or simple pull request. --fix-loop fixes and re-reviews. Complex pull request: workflow-review-changes; rationale: why-review.
日本語の概要は準備中です。原文の説明を表示しています。