The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from the web through search results and full-page scraping.
Ingestion points: Web search excerpts (data.web), tool descriptions (data.tools), and full page content via firecrawl scrape.
Boundary markers: No explicit instruction delimiters or sanitation steps are defined to prevent the agent from following instructions embedded in the retrieved web content.
Capability inventory: The agent has access to the Bash tool to execute firecrawl CLI commands.
Sanitization: None provided in the instruction set.
[EXTERNAL_DOWNLOADS]: The skill instructions and configuration refer to npx firecrawl-cli, which utilizes the Node.js package executor to download and run code from the npm registry at runtime.
[DATA_EXFILTRATION]: The skill incorporates a search-feedback mechanism that transmits search identifiers, usage metadata, and user-provided ratings back to the vendor's API. While this is documented as a credit-refund feature and includes environment-based opt-out flags, it represents a structured outbound data flow of session information.

firecrawl-search