fittings-security
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a local Node.js script 'scripts/call.mjs' to interact with remote security databases.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as vulnerability identifiers and entity names, creating an attack surface for indirect prompt injection.
- Ingestion points: Untrusted CVE IDs and person/company names enter the agent context via tool parameters.
- Boundary markers: The instructions explicitly guide the agent to treat API results as 'leads' to be verified rather than authoritative facts.
- Capability inventory: The agent can perform network requests and shell execution via the 'scripts/call.mjs' utility.
- Sanitization: The skill identifies that specific characters are rejected by upstream sources and should be filtered before API calls.
Audit Metadata