skill-release-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides administrative Python scripts (
refresh_derivation.py,release_check.py) intended for the agent to maintain package integrity and validate release manifests within the local directory. - [INDIRECT_PROMPT_INJECTION]: The pipeline includes logic for reading and analyzing external files (such as
SKILL.mdandDERIVATION.json). This ingestion surface is secured by using deterministic static analysis—specifically file hashing and regular expression matching—rather than interpolating the content into LLM prompts. - [DYNAMIC_EXECUTION]: The inclusion of a validation test suite that uses
importlib.utilto execute local scripts is a standard development practice for verification and does not pose a runtime risk to the execution environment. - [SAFE]: The
release_check.pyscript actively enhances security by scanning skill bundles for sensitive data, including AWS access keys, GitHub tokens, OpenAI API keys, and private keys, to prevent accidental exposure of credentials.
Audit Metadata