Deep Research
Agentic deep research skill using Exa.ai for search and Claude/OpenAI for synthesis.
Invocation
code
/deepresearch <topic>
Options
| Option | Description | Default |
|---|---|---|
--depth <level> | Research depth: quick (6 queries), normal (15), deep (30) | normal |
--model <provider> | LLM: claude or openai | claude |
--output <path> | Custom output path | ~/.skills/skill-deepresearch/exports/ |
--json | Also save sources as JSON | false |
--no-firecrawl | Skip deep scraping | false |
Examples
bash
# Quick overview of a topic /deepresearch "What is RAG?" --depth quick # Standard research with Claude /deepresearch "Best practices for building production ML systems" # Deep research with OpenAI, custom output /deepresearch "Compare Next.js vs Remix" --depth deep --model openai --output ./research.md
Requirements
- •
EXA_API_KEY- Required for search - •
ANTHROPIC_API_KEY- Required for Claude synthesis - •
OPENAI_API_KEY- Required for OpenAI synthesis - •
FIRECRAWL_API_KEY- Optional for deep scraping
Output
Reports are saved to ~/.skills/skill-deepresearch/exports/ with format:
- •
report-{topic-slug}-{timestamp}.md- Full research report - •
sources-{topic-slug}-{timestamp}.json- Raw sources (if --json)