ov-agentic-workflows
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill establishes security-conscious guidelines for authoring and maintaining GitHub Actions agentic workflows. It explicitly instructs the agent to minimize permissions and isolate all side effects (e.g., commenting, merging) via dedicated 'safe-output' jobs rather than granting broad permissions to the AI agent itself.
- [COMMAND_EXECUTION]: The skill utilizes the
gh aw compilecommand for workflow compilation and a Python one-liner (jsonschema.validate) for JSON schema validation. These are standard developer tools used locally for artifact generation and testing. - [EXTERNAL_DOWNLOADS]: The skill references authoritative documentation for the
gh-awframework hosted on a trusted GitHub domain (github.github.com). This is categorized as a safe reference to project-relevant documentation. - [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for managing workflows that process untrusted data (such as pull request diffs and logs). It mitigates injection risks by advising the maintenance of trigger guards, source-inspection safeguards (budgets), and explicit boundary markers, characterizing a well-controlled attack surface.
Audit Metadata