release-cherry-pick

Pass

Audited by Gen Agent Trust Hub on Oct 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes standard Git commands (git cherry-pick, git checkout, git push) and GitHub CLI commands (gh api, gh pr) to automate repository management tasks. These operations are restricted to the author's own repository (NVIDIA/Model-Optimizer) and represent expected functionality for a release engineering tool.
  • [INDIRECT_PROMPT_INJECTION]: The workflow involves fetching external pull request data (titles, numbers, and merge dates) via the GitHub API to generate PR bodies and edit labels. While this introduces an ingestion surface for external data, the risk is mitigated by the procedural nature of the skill and its focus on a specific, vendor-controlled repository.
  • [SAFE]: All external resources, including the target repository and the use of the GitHub CLI, align with the vendor's identity ('nvidia'). No suspicious patterns, such as obfuscation, credential exfiltration, or unauthorized privilege escalation, were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 2, 2026, 08:06 PM