firecrawl-map
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external site structures through the
firecrawl mapcommand. This represents a potential surface for indirect prompt injection if a targeted website provides malicious or deceptive URL paths to influence subsequent agent actions. - Ingestion points: External website structures processed by the Firecrawl CLI (SKILL.md).
- Boundary markers: None; discovery results are saved to
.firecrawl/for further processing. - Capability inventory: Bash command execution for the Firecrawl CLI tool.
- Sanitization: Standard URL mapping; no specific sanitization logic is described in the prompt template.
- [EXTERNAL_DOWNLOADS]: The skill allows the use of
npx firecrawl-cli, which downloads and executes the vendor's command-line interface. This is a legitimate tool associated with the skill author's infrastructure. - [SAFE]: The skill follows standard operational procedures for web mapping, such as saving output to a designated local directory and providing clear usage documentation. No credentials, obfuscation, or persistence mechanisms were detected.
Audit Metadata