comply
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides implementation examples that fetch binaries from external repositories. Specifically, it references the official repository of a well-known policy-as-code tool. Evidence: references/policy-as-code.md includes a GitHub Actions step to download the conftest binary from github.com/open-policy-agent/conftest.
- [COMMAND_EXECUTION]: Documentation includes instructions for system-level binary installation within a CI/CD context. Evidence: references/policy-as-code.md contains a shell command to move a binary into /usr/local/bin/ using sudo.
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface due to its core function of analyzing external codebase configurations. 1. Ingestion points: The agent reads 'control implementations' and 'evidence artifacts' from the project environment via the ASSESS workflow. 2. Boundary markers: Not explicitly defined in the prompts to segregate audited data from instructions. 3. Capability inventory: The skill generates executable policies (Rego, Kyverno) and designs audit trails. 4. Sanitization: No explicit content filtering or validation for audited artifacts is defined.
Audit Metadata