docs-researcher
Documentation Researcher Skill
Mode 1: Initialize (init)
When argument is init:
- Run the initialization script:
node ${CLAUDE_PLUGIN_ROOT}/scripts/init.js .
2. Check if `CLAUDE.md` exists (at project root):
- If exists: Read it and append the knowledge base section if not already present
- If not exists: Create it with the knowledge base section
**Append/create this section in CLAUDE.md:**
```markdown
## Knowledge Base
This project uses `.claude/skills/project-knowledge-base/` for documentation.
**Before coding unfamiliar patterns:**
1. Load the `project-knowledge-base` skill to see available references
2. If topic not covered, use `/docs-researcher <technology> <topic> for <context>`
Always consult the knowledge base when clarification is needed.
Example:
User: /docs-researcher init
Actions:
1. Bash(node ${CLAUDE_PLUGIN_ROOT}/scripts/init.js .)
2. Read CLAUDE.md (if exists, at project root)
3. Write updated CLAUDE.md with knowledge base section
Response: Report what was created/updated
Mode 2: Research
When argument is a research request (not "init"), invoke the docs-researcher agent.
Task(
subagent_type: "docs-researcher",
description: "Research documentation",
prompt: "{user's research request}"
)
Agent will:
- Search the web for documentation
- Fetch and extract relevant content
- Save to
.claude/skills/project-knowledge-base/references/ - Update the SKILL.md index
Example:
User: /docs-researcher TanStack Router beforeLoad for authentication guards
Action: Task(subagent_type: "docs-researcher", prompt: "TanStack Router beforeLoad for authentication guards")
Response: Agent's research results
More from amoscicki/aromatt
semantic-search
Semantic codebase search with pgvector. Finds code by MEANING. **USE THIS FIRST** before grep/glob. Returns file paths + byte ranges for precise reads. Multi-project support with auto-discovery — no manual setup needed.
14research-methodology
This skill should be used when docs-researcher agent needs guidance on "how to search documentation", "WebSearch query patterns", "filtering search results", "documentation research strategy", or "creating knowledge files". Provides systematic methodology for effective technical documentation research.
11payload-cms
Query and mutate Payload CMS collections via persistent local server. Use for reading/writing data in dev or test databases.
10gapi
Run local Node CLIs for Google Tag Manager and Google Analytics Admin APIs (OAuth2). Use for GTM containers/tags/triggers or GA properties/data-streams.
9