update-doc

Use when user says 'update docs', 'sync documentation', 'refresh docs', or after code changes need documenting. Updates llmdoc system based on recent code changes.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "update-doc" with this command: npx skills add tokenrollai/cc-plugin/tokenrollai-cc-plugin-update-doc

/update-doc

This skill updates the project's llmdoc documentation to reflect recent code changes.

Pre-fetched Context

  • Llmdoc structure: !find llmdoc -name "*.md" 2>/dev/null | head -50
  • Recent changes (3 commits): !git diff HEAD~3..HEAD --stat 2>/dev/null | head -30
  • Affected files: !git diff HEAD~3..HEAD --name-only 2>/dev/null | head -30
  • Llmdoc index: !cat llmdoc/index.md 2>/dev/null | head -50

Actions

  1. Step 1: Analyze Changes

    • If $ARGUMENTS is provided, use it as the description of what changed.
    • Otherwise, analyze the pre-fetched git diff to understand what changed.
  2. Step 2: Identify Impacted Concepts

    • Map changed files to llmdoc concepts:
      • Config files (.eslintrc, etc.) → reference/coding-conventions.md
      • Auth files → relevant architecture docs
      • New features → may need new docs
    • Create a list of impacted documents.
  3. Step 3: Update Documents

    • For each impacted document, use recorder agent with this prompt:
      Task: Update documentation for <concept_name>.
      Changes: <relevant git diff summary>
      Mode: content-only
      Principle: Use fewest words necessary.
      
  4. Step 4: Synchronize Index

    • After all updates complete, invoke a single recorder agent to:
      • Re-scan /llmdoc directory
      • Ensure index.md is consistent and up-to-date
      • Mode: full
  5. Step 5: Report

    • Summarize all documents created/updated/deleted.

Update Principles

  • Minimality: Use fewest words necessary
  • Accuracy: Based on actual code, not assumptions
  • No Code Blocks: Reference with path/file.ext:line format
  • LLM-Friendly: Write for machine consumption

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

doc-workflow

No summary provided by upstream source.

Repository SourceNeeds Review
General

read-doc

No summary provided by upstream source.

Repository SourceNeeds Review
General

commit

No summary provided by upstream source.

Repository SourceNeeds Review