本文へ移動
cccskills
無料GitHub で公開

doc

SOP for structured documentation lifecycle. Trigger when: - Designing, writing, reviewing, or auditing large-scale documentation structures. - Navigating phases: Audit, Plan, Draft, Review, Verify. - Prompt contains: /doc, doc workflow, audit, draft, review, verify, documentation debt.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md7.4 KB

SKILL.md(原文)

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。

DOC Protocol

Audit → Plan → Draft → Review → Verify

You are a Documentation Engine. Your goal is to produce or improve documentation through a structured lifecycle that ensures every document is purposeful, accurate, and maintainable.

DOC applies the documentation rule (skills/documentation/SKILL.md) systematically. The rule provides the principles; this workflow provides the process.

Routing: DOC is the structured workflow for deliberate documentation work. The documentation skill is the writing-quality authority and governs all text production; DOC wraps it in a lifecycle. If the scope is ambiguous (e.g., a single inline update vs. a structural audit), set UNCERTAINTY > 0.0 and surface the question before proceeding to PLAN.


Scope

[!IMPORTANT] DOC is for deliberate documentation work — writing new guides, restructuring docs, running documentation audits, or remedying documentation debt. It is NOT for incidental text (commit messages, code comments, chat responses) — the documentation skill's Section 1 governs those automatically.

When to use /doc:

  • Writing or rewriting a README, guide, tutorial, or reference page
  • Auditing a project's documentation for completeness and accuracy
  • Systematically addressing documentation debt
  • Restructuring documentation architecture (reorganizing, splitting, merging)

When NOT to use /doc:

  • Quick inline doc updates triggered by a code change — just apply the documentation skill directly
  • Writing a plan, sketch, or ADR — those have their own workflows

Grammar

# 1. STATE METADATA
STATUS: [AUDIT | PLAN | DRAFT | REVIEW | VERIFY]
UNCERTAINTY: [0.0-1.0] # Residual entropy. Must be 0.0 to proceed.

# 2. CONTEXT
CTX:
  SCOPE: "What documentation work is being done"
  AUDIENCE: "Who will read the output"
  QUADRANT: [TUTORIAL | HOW_TO | REFERENCE | EXPLANATION | MIXED]
  # MIXED only for multi-document work (e.g., audit). Single documents must declare one quadrant.

# 3. AUDIT FINDINGS (Populated in AUDIT)
AUDIT:
  EXISTING:
    - DOC: "path/to/existing/doc.md"
      QUADRANT: "What type it currently is"
      STATUS: [CURRENT | STALE | MISLEADING | ORPHANED | MISSING]
      NOTES: "What needs attention"
  DEBT:
    - ITEM: "Description of documentation debt"
      SEVERITY: [LOW | MEDIUM | HIGH]
      ACTION: "What to do about it"

# 4. DOC PLAN (Populated in PLAN)
DOC_PLAN:
  DELIVERABLES:
    - FILE: "path/to/new/or/updated/doc.md"
      QUADRANT: "TUTORIAL | HOW_TO | REFERENCE | EXPLANATION"
      AUDIENCE: "Who this is for"
      ACTION: [CREATE | REWRITE | UPDATE | DELETE]
      SUMMARY: "What this document will contain"

# 5. REVIEW CHECKLIST (Populated in REVIEW)
REVIEW:
  DOCUMENTATION_STANDARDS:
    ANSWER_FIRST: [PASS | FAIL]
    ACTIVE_VOICE: [PASS | FAIL]
    TERMINOLOGY_LOCK: [PASS | FAIL]
    SELF_CONSISTENCY: [PASS | FAIL]
    SCOPE_ECONOMY: [PASS | FAIL]
  QUADRANT_FIT: [PASS | FAIL]
  AUDIENCE_MATCH: [PASS | FAIL]
  NOTES: "Issues found during review"

State Transitions

AUDIT ──→ PLAN    (once existing docs catalogued and debt identified)
      └─→ HALT    (if no documentation work is needed — a valid outcome)

PLAN ──→ DRAFT   (on human approval of doc plan)
     └─→ AUDIT   (if planning reveals undiscovered existing docs)

DRAFT ──→ REVIEW (once document text is written)
      └─→ PLAN   (if drafting reveals scope was wrong)

REVIEW ──→ VERIFY (once review checklist passes)
       └─→ DRAFT (if review finds failures requiring redraft)

VERIFY ──→ DONE  (on human approval)
       └─→ REVIEW (if verification reveals issues)

State Definitions

AUDIT: Catalogue existing documentation. For each document, identify its Divio quadrant, currency status, and any debt. This step may be skipped if the scope is a single new document with no existing context.

PLAN: Define what will be written, rewritten, or deleted. Each deliverable declares its quadrant, audience, and action. Present for human approval before drafting.

DRAFT: Write the documentation. Apply all documentation skill principles (Section 1 + Section 2). Produce complete text, not outlines or placeholders.

REVIEW: Self-critique against the documentation skill's principles. Run through the review checklist honestly. If anything fails, return to DRAFT — do not rationalize a pass.

VERIFY: Present the finished documentation for human review. The human confirms accuracy, completeness, and fit. If verification fails, return to REVIEW with specific feedback.


Prime Directives

  1. QUADRANT_FIRST: Identify the document type before drafting. A document without a declared quadrant is a document without a purpose.

  2. AUDIENCE_LOCK: State who the reader is. If you can't name the audience, you can't write for them.

  3. ONE_TYPE_PER_DOCUMENT: Do not blend quadrants within a single document. A tutorial that becomes a reference page serves neither audience. If mixed content is needed, produce separate documents.

  4. DEBT_VISIBILITY: Documentation debt discovered during AUDIT must be recorded, even if it's out of scope for the current /doc invocation. Flag it for future work.

  5. HONEST_REVIEW: The REVIEW phase is self-adversarial. Apply documentation checks mechanically — if active voice fails, it fails. Do not hand-wave compliance.

  6. HUMANIZER_INTEGRATION: After REVIEW passes, run the output through /humanizer if the document is reader-facing (README, guide, tutorial). Skip for internal references and specifications. The documentation skill governs structure and clarity; the humanizer governs natural voice.

Protocol Violations (FORBIDDEN)

ViolationWhy It's Wrong
Drafting without declaring quadrant and audiencePurposeless writing; serves no reader well
Blending Tutorial and Reference in one documentFails both audiences — tutorial readers get lost, reference users can't scan
REVIEW that passes everything without commentSelf-review is adversarial; a clean pass requires justification
Skipping AUDIT for multi-document workYou'll duplicate or contradict existing docs
Marking stale docs as "current" to avoid reworkDocumentation debt hiding; compounds over time

MANDATORY HALT Points

You MUST stop and await human input at:

  1. AUDIT → PLAN transition: Human must approve the documentation plan
  2. VERIFY: Human must confirm the final output
  3. HALT from AUDIT: If audit reveals no work is needed, confirm with human

Response Format

  • AUDIT: YAML block with AUDIT findings
  • PLAN: YAML block with DOC_PLAN deliverables
  • DRAFT: YAML block + full document text
  • REVIEW: YAML block with REVIEW checklist + notes
  • VERIFY: Final document text for human approval

レビュー

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。

同じリポジトリのスキル

概要と使いどころ

ai-audit

無料

SOP for auditing AI-generated code. Trigger when: - Reviewing, refactoring, or cleaning up AI-generated code to prevent regressions or hallucinated APIs. - Prompt contains: /ai-audit, code audit, AI cleanup, common flaws.

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。

nrdxp/predicate102026年9月1日 更新

api-audit

無料

Protocol for auditing API surface coherence and type safety. Trigger when: - Evaluating API designs, interface type safety, or design elegance. - Prompt contains: /api-audit, API surface, API coherence, type safety.

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。

nrdxp/predicate102026年9月1日 更新

boundary

無料

Normative sufficiency conditions for Initial Boundary Conditions (IBCs) and the SOP for the cheap-tier boundary refinement loop (/boundary). Trigger when: - Crafting, auditing, or refining a prompt/IBC destined for an expensive (architect-class) model or an autonomous worker dispatch. - Evaluating whether a task frame is sufficient to bound an agent walk. - Prompt contains: /boundary, IBC, initial boundary condition, boundary contract, sufficiency conditions, worker prompt, prompt refinement.

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。

nrdxp/predicate102026年9月1日 更新

campaign

無料

SOP for the architect-tier campaign workflow (/campaign): exhaustive survey, mitigation planning, tiered orchestration, and reconciliation. Trigger when: - Running a multi-workstream initiative where an expensive architect-tier council surveys, plans, emits worker prompts, and judges landed work. - Conducting production-readiness assessments that fan out into autonomous mitigation dispatches across model tiers. - Prompt contains: /campaign, campaign workflow, survey, orchestrate, reconcile, premise freshness, tier routing, worker IBC, scratch.

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。

nrdxp/predicate102026年9月1日 更新

chronicle

無料

Maintain and update the persistent project chronicle (docs/chronicle.md). Trigger when: - The human requests a history summary or chronicle update. - Starting work on a new codebase and needing context on its evolution. - Prompt contains keywords: /chronicle, chronicle, project history, git log summary, history summary.

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。

nrdxp/predicate102026年9月1日 更新

Rules, conventions, and constraints for formatting git commit messages and committing at logical boundaries. Trigger when: - Drafting, revising, or validating git commit messages. - Pausing at commit boundaries under the CORE or CONTINUE workflows. - Evaluating whether a changeset should be split into multiple commits. - Prompt contains keywords: commit message, git commit, conventional commits, commit hygiene, commit guidelines, logical boundary, spaghetti diff, atomic commit, commit boundary.

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。

nrdxp/predicate102026年9月1日 更新

nrdxp のスキルをすべて見る

このスキルの問題を報告する