skills/nvidia/nemo-relay/maintain-ci/Gen Agent Trust Hub

maintain-ci

Pass

Audited by Gen Agent Trust Hub on Oct 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill encourages security hardening for CI/CD pipelines, specifically recommending full commit SHAs for third-party actions and implementing the principle of least privilege for GitHub job permissions.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub Actions workflow files, which could theoretically contain malicious instructions.
  • Ingestion points: Reads and reviews .github/workflows/*.yml and .github/workflows/*.yaml files.
  • Boundary markers: None explicitly defined in the instructions to separate workflow configuration data from agent instructions.
  • Capability inventory: The skill uses ruby for YAML validation, uv for pre-commit hooks, and ripgrep for workflow graph inspection. It possesses file-system access to read and edit CI configurations.
  • Sanitization: Relies on standard YAML parsing and existing pre-commit hooks; no specific prompt-level sanitization is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 3, 2026, 05:19 PM