pentest-secrets-exposure
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, including source code, git history, and web application responses, which creates a surface for indirect prompt injection attacks.
- Ingestion points: Untrusted data enters the agent context through
repo_path(TruffleHog/GitLeaks/Semgrep),url(SecretFinder/nuclei/ffuf/error_trigger), and analyzed files such as.envor JavaScript bundles. - Boundary markers: The instructions do not define specific delimiters or instructions to prevent the agent from obeying instructions embedded within the analyzed target data.
- Capability inventory: The skill utilizes network operations (
nuclei,curl,credential_validator), file system access for repository scanning, and command execution via various security tools. - Sanitization: No explicit sanitization, filtering, or validation of the content retrieved from targets is described before it is processed by the agent.
Audit Metadata