The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the web via search results and page scraping. This data is ingested into the agent's context, creating a surface for indirect prompt injection.
Ingestion points: Results from firecrawl search and firecrawl scrape commands, which are often stored in .firecrawl/ JSON files or output directly to the context.
Boundary markers: The skill does not define explicit delimiters or instructions for the agent to ignore commands within retrieved web data.
Capability inventory: The agent is granted Bash tool access, which can be used for file operations, network requests, and executing CLI tools.
Sanitization: No specific sanitization or filtering of scraped web content is described before it reaches the agent.
[EXTERNAL_DOWNLOADS]: The documentation references the use of npx firecrawl-cli, which triggers the download of the Firecrawl CLI package from the public npm registry at runtime.
[COMMAND_EXECUTION]: The skill relies on the Bash tool to execute firecrawl CLI commands for searching, listing tool contracts, scraping data, and sending feedback telemetry.

firecrawl-search