The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill navigates and extracts data from external, untrusted websites. This content could contain malicious instructions designed to influence the agent's behavior during the extraction process.
Ingestion points: Website content accessed via the firecrawl agent command in SKILL.md.
Boundary markers: None specified in the instructions to delimit untrusted web content.
Capability inventory: Execution of shell commands via Bash(firecrawl *) and Bash(npx firecrawl-cli *).
Sanitization: Not explicitly implemented in the skill; relies on the underlying tool's internal guardrails.
[COMMAND_EXECUTION]: The skill executes firecrawl and npx firecrawl-cli commands. These tools interact with the local file system to create a .firecrawl/ directory and write extraction results to JSON files.
[EXTERNAL_DOWNLOADS]: The skill uses npx to download and execute the firecrawl-cli package. As firecrawl-cli is an official vendor resource for this skill, it is documented as a standard functional dependency.

firecrawl-agent