actions-pinning
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted GitHub Action manifest files, representing an attack surface for indirect prompt injection.
- Ingestion points: The skill inventories files like
.github/workflows/*.ymlandaction.ymlusing the bundled scripts/pin_check.py script. - Boundary markers: Not present; data is ingested as plain text/YAML.
- Capability inventory: The skill can execute shell commands (python3, gh api), perform file writes to update manifests, and access the GitHub API.
- Sanitization: The scanner uses strict regular expressions to isolate action references, reducing the risk of processing malicious instructions embedded in workflow comments or unrelated fields.
- [COMMAND_EXECUTION]: The skill executes the bundled scripts/pin_check.py script and utilizes the GitHub CLI (gh api) to resolve and verify repository data. These executions are legitimate functions for the skill's primary purpose of maintaining secure workflows.
- [EXTERNAL_DOWNLOADS]: The skill fetches repository and commit metadata from official GitHub API endpoints. These requests target a well-known, trusted service and do not involve the download or execution of untrusted third-party code.
Audit Metadata