firecrawl-download

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by downloading content from external websites to the local filesystem. 1. Ingestion points: External web content fetched via firecrawl download commands. 2. Boundary markers: None specified to delimit untrusted content or warn the agent about embedded instructions. 3. Capability inventory: Bash command execution for site mapping and writing files to the local filesystem. 4. Sanitization: No explicit mention of sanitizing or filtering external content before storage.
  • [COMMAND_EXECUTION]: The skill invokes the firecrawl CLI through the Bash tool. It specifically advises using the -y flag to bypass confirmation prompts, which increases the agent's autonomy during automated site downloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 02:12 AM