Research Skill

---

description: "Research a topic using autonomous web search and AI synthesis" argument-hint: " [--depth quick|medium|deep]"

Research Skill

Research a topic using web search and specialist agents. Results are synthesized by AI and stored for future reference.

Arguments

The query to research, optionally followed by depth:

• --depth quick - Fast lookup (~10s) for simple facts
• --depth medium - Standard research (~30s) for how-to questions (default)
• --depth deep - Comprehensive analysis (~60s) for comparisons

How to Execute

1. Parse the query from: $ARGUMENTS
2. Extract depth if specified (default: medium)
3. Call the research service API:

curl -X POST http://localhost:3200/api/research \
  -H "Content-Type: application/json" \
  -d '{"query": "", "depth": "", "trigger": "manual"}'

1. Poll for results at: GET http://localhost:3200/api/research/
2. Present the findings to the user

Alternative: Direct Execution

For immediate results without polling, use the execute endpoint:

curl -X POST http://localhost:3200/api/research/execute \
  -H "Content-Type: application/json" \
  -d '{"query": "", "depth": ""}'

This returns results directly (blocks until complete). Timeouts: quick=90s, medium=180s, deep=300s.

Examples

• /research What is HTMX - Quick fact lookup
• /research How to implement rate limiting in FastAPI - Medium depth how-to
• /research Rust vs Go for CLI tools --depth deep - Deep comparison

Service Status

If the service isn't running, inform the user to start it:

cd /home/bigphoot/Desktop/Projects/claude-research-team && npm run start

Dashboard available at: http://localhost:3200