Searches, maps, and scrapes web pages through the Firecrawl CLI with cached, context-limited retrieval.
npx skills add https://github.com/nejcm/agents --skill firecrawl-nejcm
Please help me install this Agent Skill. Skill: firecrawl Source: https://github.com/nejcm/agents/tree/main/raw/all/skills/firecrawl Command: npx skills add https://github.com/nejcm/agents --skill firecrawl-nejcm
search returns at most five result records and map returns up to 20 matching URLs from a known documentation site, without scraping pages. - Selective page retrieval: summary and read write page content to local cache files so you can grep and read only relevant ranges instead of loading entire pages. - 24-hour caching: Repeated queries reuse cached results, with optional refresh via environment variables. - Use Case: When asked how a specific API works, search for the official docs, map the documentation site, then read at most two authoritative pages to answer with accurate details. ## Quick Start Ask the agent to research a topic on the web using the firecrawl skill, for example to find and summarize the official documentation for a specific API endpoint.High-intent search queries and answers about installing and using this skill.
Run the wrapper script with the search command followed by your query. It calls the Firecrawl CLI and returns at most five metadata-only results with title, URL, and description, without scraping any pages.
Use the map command with a base URL and query to list up to 20 matching URLs, then call summary or read on at most two pages. Results are written to local cache files that you grep and read selectively.
It requires Node.js to run the wrapper script and the global firecrawl-cli package installed via npm install -g firecrawl-cli@1.19.26. Authentication uses the FIRECRAWL_API_KEY environment variable or Firecrawl's stored login.
The global firecrawl CLI is missing or not on the PATH. Install it with npm install -g firecrawl-cli@1.19.26; the wrapper does not support npx or Firecrawl's full agent-skill bundle as substitutes.
Results are cached for 24 hours by default in the OS temp directory. Set FIRECRAWL_AGENT_CACHE_TTL_MS to change the lifetime, FIRECRAWL_AGENT_REFRESH=1 to bypass a cached result once, or run the clear command to empty the cache.
wanxingaiSearch local notes and distinguish evidence from unsupported claims.
gagan114662Search the web, evaluate sources, and synthesize findings with cited URLs.
ai-service-incubatorVerify cosmetic product information against official sources and generate structured reports.
MadeAgentsSummarize URLs, transcripts, YouTube links, and local files into condensed output.