firecrawl-build-scrape
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest content from arbitrary external URLs which are processed for downstream LLM use.
- Ingestion points: Content is retrieved from user-provided URLs using the Firecrawl
/scrapeAPI (SKILL.md). - Boundary markers: The skill does not explicitly define prompt delimiters for the scraped content, though it mentions setting an extraction contract.
- Capability inventory: The skill allows fetching markdown, HTML, links, and metadata from external web pages.
- Sanitization: The skill recommends using
onlyMainContentto filter out non-essential page elements, which reduces the surface for embedded malicious instructions. - [EXTERNAL_DOWNLOADS]: The skill provides references and links to official documentation and source code repositories for the Firecrawl service on the
docs.firecrawl.devandgithub.comdomains. - [CREDENTIALS_UNSAFE]: The skill defines a
FIRECRAWL_API_KEYas a required input field. This follows standard practices for secure credential management by allowing keys to be provided via the environment or inputs rather than hardcoding them in the skill text.
Audit Metadata