firecrawl

Pass

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS] (LOW): The skill requires the installation of the firecrawl-cli package from the npm registry.
  • Evidence: Found in rules/install.md: npm install -g firecrawl-cli@1.4.1.
  • Note: While the author is not on the predefined trusted list, the package is a versioned release from a standard public registry.
  • [PROMPT_INJECTION] (LOW): The skill is susceptible to Indirect Prompt Injection because it retrieves and processes untrusted data from the internet.
  • Ingestion points: SKILL.md (via search, scrape, crawl, and agent commands).
  • Boundary markers: Present. rules/security.md mandates using the -o flag to write output to local files in the .firecrawl/ directory rather than injecting raw content directly into the prompt context.
  • Capability inventory: The skill allows shell execution of the firecrawl CLI and utilizes the agent command for autonomous extraction (SKILL.md).
  • Sanitization: Present. rules/security.md instructs the agent to use URL quoting in shell commands and perform incremental reads (e.g., head, grep) to limit the amount of untrusted content loaded into the LLM context window at once.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 18, 2026, 08:44 PM