firecrawl-workflows
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from the web via Firecrawl tools to produce research reports and other deliverables.
- Ingestion points: Web evidence is collected through the Firecrawl CLI or equivalent tool surfaces as described in the 'Default Process' section of SKILL.md.
- Boundary markers: The instructions mandate citing source evidence and including an evidence base, though they do not specify explicit LLM-level delimiters for the external content.
- Capability inventory: The skill uses the Firecrawl CLI, parallel task runners, and sub-agents to process data.
- Sanitization: There are no explicit instructions for sanitizing or escaping the web content before synthesis.
- [SAFE]: The skill references established repositories from a trusted organization (Anthropics) as guidelines for workflow authoring in the 'references/workflow-authoring.md' file.
- [SAFE]: The skill manages its API key (FIRECRAWL_API_KEY) as a required user input rather than hardcoding credentials, which is a standard security practice.
Audit Metadata