STOP - Before using this skill for ANY Claude Code documentation query:

IF YOU ARE THE MAIN AGENT, you MUST invoke BOTH sources in the same message:

1. This skill (

   docs-management

   ) - local cache, token-efficient

2. claude-code-guide

   subagent - live web search

text

[Skill tool: docs-management]
"Find documentation about {topic}"

[Task tool: claude-code-guide] (SAME MESSAGE - USE THIS EXACT PROMPT)
"First WebFetch https://code.claude.com/docs/en/claude_code_docs_map.md to find
relevant doc pages about {topic}. Then WebFetch those specific pages. Use WebSearch
only if needed for additional context. Do NOT use Skill tool (not available).
Return key findings with source URLs."

⚠️ CRITICAL: claude-code-guide does NOT have Skill tool access. Always prompt it to use WebSearch/WebFetch explicitly. If you see "No such tool: Skill" error, you prompted it wrong.

This is AUTOMATIC. Do NOT wait for user to ask for it.

IF YOU ARE A SUBAGENT: Note in your response that main agent should also query

claude-code-guide

.

cd

&&

1. ✅ ALWAYS use helper scripts (recommended - they handle path resolution automatically)
2. ✅ Use absolute path resolution (if not using helper script)
3. ✅ Use separate commands (never

   cd

   with

   &&

   )

• ❌ Chain

  cd

  with

  &&

  :

  cd <relative-path> && python <script>

  causes path doubling
• ❌ Assume current directory
• ❌ Use relative paths when current dir is inside skill directory

• /claude-ecosystem:scrape-docs

  - Scrape all configured Claude documentation sources, then refresh index and validate

• /claude-ecosystem:refresh-docs

  - Refresh the local index and metadata without scraping from remote sources

• /claude-ecosystem:validate-docs

  - Validate the index and references for consistency and drift without scraping

• Canonical storage (encapsulated in skill) - Single source of truth for official docs
• Subsection extraction - Token-optimized extracts for skills (60-90% savings)
• Drift detection - Hash-based validation against upstream sources
• Sync workflows - Quarterly maintenance automation
• Documentation discovery - Keyword-based search and doc_id resolution
• Index management - Metadata, keywords, tags, aliases for resilient references

search

search skills progressive

query

query "how to create skills"

content

content code-claude-com-docs-en-skills

content --section

content <doc_id> --section "metadata"

resolve

resolve code-claude-com-docs-en-skills

related

related code-claude-com-docs-en-skills

category

category api

tag

tag skills

search

doc_id

content

• search

  and

  query

  accept flexible keywords (no full doc_id needed)

• content

  ,

  resolve

  , and

  related

  require full doc_id (e.g.,

  code-claude-com-docs-en-skills

  )

• Scraping documentation - Fetching docs from sitemaps or docs maps
• Finding documentation - Searching for docs by keywords, category, or natural language
• Resolving doc references - Converting doc_id to file paths
• Managing index metadata - Adding keywords, tags, aliases, updating metadata
• Rebuilding index - Regenerating index from filesystem (handles renames/moves)

CRITICAL: This section is the authoritative source for Claude Code documentation access patterns.

• When main agent loads this skill → Main agent must also invoke

  claude-code-guide

  in parallel
• When subagent loads this skill → Subagent should note in response that main agent should invoke

  claude-code-guide

  for live coverage
• When command invokes this skill → Same as main agent (commands run in main context)

• You MUST invoke both this skill AND

  claude-code-guide

  subagent in the SAME message
• This is the only way to get comprehensive Claude Code documentation coverage

• You CANNOT invoke

  claude-code-guide

  yourself (architectural constraint)
• Include in your response: "For comprehensive coverage, main agent should also query claude-code-guide subagent"
• Proceed with local cache results

1. Invoke

   docs-management

   skill (local cache)
2. Spawn

   claude-code-guide

   subagent (live web) in the same message

• Claude Code features (hooks, skills, memory, MCP, plugins, settings, etc.)
• How to do something in Claude Code
• Claude Code configuration or troubleshooting
• Any topic in the Claude Code documentation

• Combine results from both sources
• Local cache provides category context and token efficiency
• Live search provides currency and gap-filling
• Note any discrepancies between sources

• Plugin not installed: Use

  claude-code-guide

  only
• Live search unavailable: This skill provides complete local cache
• Both sources disagree: Flag discrepancy, prefer official docs, investigate

docs-management

claude-code-issue-researcher

claude-code-guide

• "error", "bug", "broken", "not working", "fails", "crash"
• "debug", "troubleshoot", "workaround", "issue", "problem"
• "doesn't work", "won't work", "can't", "isn't firing"

Search GitHub issues in anthropics/claude-code for: [ERROR/PROBLEM DESCRIPTION].
Check both open and closed issues. Report issue numbers, status, and any workarounds.

• subagent_type

  : "general-purpose"

• description

  : Short 3-5 word description

• prompt

  : Full task description with execution instructions

• ✅ Run directly:

  python .claude/skills/docs-management/scripts/core/scrape_all_sources.py --parallel --skip-existing

• ✅ Streaming logs: Scripts emit progress naturally via stdout
• ✅ Wait for completion: Scripts exit when done with exit code
• ❌ NEVER use

  run_in_background=true

  : Scripts are designed for foreground execution
• ❌ NEVER poll output: Streaming logs appear automatically, no BashOutput polling needed
• ❌ NEVER use background jobs: No

  &

  , no

  nohup

  , no background process management

1. Task agent reports "Done (X tool uses · Y tokens · Z time)"
2. ✅ READ the agent's message containing its final report
3. ✅ SUMMARIZE the agent's findings to the user
4. ❌ NEVER use BashOutput to "check the real results"
5. ❌ NEVER poll or verify what the agent already reported

• ✅ Report script errors: Exit codes, exceptions, error messages
• ✅ Report warnings: Deprecation warnings, import issues, configuration problems
• ✅ Report unexpected output: 404s, timeouts, validation failures
• ✅ Include context: What was being executed when the error occurred

1. docs.claude.com and code.claude.com: These domains serve .md URLs successfully
2. anthropic.com: These domains do NOT serve .md URLs (404 expected, configured with

   try_markdown: false

   )

• Python 3.14 works by default for scraping
• Python 3.13 required for spaCy operations (keyword extraction, metadata extraction)

undefined

search

1. Index search - Fast keyword matching against document metadata (title, description, keywords, tags)
2. Content search - Grep through actual file contents for matches not in index

• [SUBSECTION]

  - Match found in a specific document section (click to extract)

• [CONTENT]

  - Match found in file content (not in metadata index)
• No marker - Standard index match

• Use

  --fast

  for index-only search (faster, but may miss content-only matches)
• Use

  --separate

  to display index and content matches in separate sections

• Title matches (highest weight, 14 points)
• Description matches (10 points)
• Keyword field matches (8-10 points, with variant detection)
• Tag matches (5-7 points)
• Coverage multipliers (1.5-2.0x for multi-term matches)
• Domain weighting (Claude Code docs prioritized over general Anthropic docs)

--verbose

• config/defaults.yaml

  - Central configuration file with all default values

• config/config_registry.py

  - Canonical configuration system with environment variable support

• references/sources.json

  - Documentation sources configuration (required for scraping)

config/defaults.yaml

paths

CLAUDE_DOCS_<SECTION>_<KEY>

• Adding New Documentation Source - Onboarding new docs from sitemaps or docs maps
• Scraping Multiple Sources - Validation checkpoints to prevent wasted time/tokens
• Token-Optimized Subsection Retrieval - Workflow for extracting subsections instead of full documents

• Windows:

  winget install --id Python.Python.3.13 -e --source winget

• macOS:

  brew install python@3.13

• Linux:

  sudo apt install python3.13

• Update workflow: Scrape → Validate → Review → Commit → Version bump → Push
• Version bumps: Patch for doc refresh, Minor for new sources/features, Major for breaking changes
• Testing: Run

  manage_index.py verify

  and test search before pushing

1. Set

   OFFICIAL_DOCS_DEV_ROOT

   in your terminal
2. Run scripts - output goes to dev repo
3. Track changes:

   git diff canonical/

4. Commit and push when ready

• index.yaml (in canonical storage) - Complete registry of docs and extracts
• references/parallelization-strategy.md - Parallelization decision trees
• references/plugin-maintenance.md - Plugin update and publishing workflows
• DEPENDENCIES.md - Script dependency graph and execution order

• v1.19.0 (2025-11-25): Determinism fixes and observability enhancements
• v1.18.0 (2025-11-18): CRITICAL FIX: Enforce foreground execution pattern
• v1.17.1 (2025-11-18): Fix Task tool invocation syntax
• v1.17.0 (2025-11-18): Critical workflow execution guidance + error reporting
• v1.16.0 (2025-11-17): Comprehensive search quality audit & fixes (100% pass rate)
• v1.15.0 (2025-11-17): Critical bug fix + domain prioritization
• v1.14.0 (2025-11-17): Comprehensive skill audit & validation (A+ grade)