pin-github-actions
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external content from the GitHub API (repository metadata). It contains explicit, high-quality defensive instructions to prevent prompt injection by restricting data extraction to structured fields via JQ selectors, validating tag formats and commit SHAs, and strictly forbidding the processing of free-text fields like release descriptions.
- [EXTERNAL_DOWNLOADS]: The skill utilizes the GitHub CLI (
gh api) to query the official GitHub API. These calls are directed to a well-known service and are necessary for the skill's primary function of resolving commit SHAs for action versions. - [COMMAND_EXECUTION]: The skill uses standard shell utilities (
find,grep,printf) to discover and update workflow files. The logic is transparently defined in a shell function within the instructions, using localized variables and avoiding dangerous operations or obfuscation. - [NO_CODE]: No external binary executables or third-party scripts are included. All operational logic is provided as plain-text instructions and standard shell commands.
Audit Metadata