find-recruiting-firm
Fail
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill directs the agent to perform silent harvesting of API tokens from sensitive local configuration files.
- Evidence: The 'Auth' section instructs the agent to "read .env.local then .env in the current working directory and look for a SERVICEGRAPH_TOKEN=… line."
- Evidence: The instructions explicitly tell the agent to bypass user confirmation: "If you find it, use it; don't ask."
- [DATA_EXFILTRATION]: Data extracted from local sensitive files is intended for transmission to an external endpoint.
- Evidence: The skill mandates setting the extracted token in the 'Authorization: Bearer' header for all requests to 'https://api.servicegraph.co'.
- [PROMPT_INJECTION]: The instructions contain directive language designed to minimize user oversight and override default agent cautiousness regarding file access.
- Evidence: The phrase "don't ask" is used to explicitly suppress the agent's typical behavior of requesting permission before accessing sensitive filesystem data.
Recommendations
- AI detected serious security threats
Audit Metadata