firecrawl-knowledge-ingest
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted web pages and documentation portals, which introduces a potential surface for indirect prompt injection if the ingested data contains malicious instructions targeting the LLM.
- Ingestion points: Public or authenticated third-party documentation portals and knowledge bases fetched via the Firecrawl browser workflow (
SKILL.md). - Boundary markers: Absent. The instructions do not specify any delimiters or safety guidelines directing the agent to disregard instructions contained within the scraped web text.
- Capability inventory: None. The skill defines natural language workflows and output templates but does not execute local commands, write files, or invoke code execution tools within the provided file.
- Sanitization: Absent. Content is transformed into markdown and structured JSON without explicit filtering or validation of the text body.
Audit Metadata