The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (MEDIUM): The skill requires the global installation of firecrawl-cli@1.4.1 via npm. While npm is a standard registry, this package and its author do not appear on the provided list of trusted organizations, making it an unverifiable dependency.
[PROMPT_INJECTION] (LOW): High surface area for indirect prompt injection (Category 8).
Ingestion points: Untrusted data enters the agent context via the scrape, search, crawl, map, and browser commands, which fetch content from arbitrary external URLs.
Boundary markers: Absent. The skill does not instruct the agent to use delimiters or ignore instructions found within the scraped markdown output.
Capability inventory: The agent has capabilities to execute shell commands (firecrawl), perform network requests (via the CLI), and write files to the local directory.
Sanitization: No sanitization or escaping of scraped content is performed before it is presented to the LLM.
[COMMAND_EXECUTION] (SAFE): The skill relies on shell command execution for its primary purpose. It follows best practices by advising the user to fix npm prefixes instead of using sudo for installation.

firecrawl