The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches data from the official Firecrawl API (api.firecrawl.dev), which is a well-known service for web scraping.
[COMMAND_EXECUTION]: Executes shell commands including curl for API requests, jq for parsing JSON data, and platform-specific diagnostic tools like zero doctor. It also involves local file system operations, such as writing request configuration to /tmp/firecrawl_request.json and saving scraped output to markdown files.
[DATA_EXFILTRATION]: Transmits user-provided URLs and natural language instructions (prompts for AI extraction) to the external Firecrawl API. This behavior is inherent to the scraping and data extraction features of the service.
[PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection because it ingests and processes content from arbitrary external websites that could contain malicious instructions.
Ingestion points: Web content retrieved from various domains via the Firecrawl API (SKILL.md).
Boundary markers: The instructions do not include specific delimiters or warnings to the agent to prevent it from executing instructions found within the scraped content.
Capability inventory: The environment allows network operations via curl and file-writing capabilities (SKILL.md).
Sanitization: There are no mechanisms described for sanitizing or filtering the scraped content before it is processed by the agent.

firecrawl