run-release

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell commands using git for version control, gh for GitHub platform interactions, and package managers pnpm and bun to run format checks and unit tests.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes external data from the repository to generate release notes.
  • Ingestion points: Untrusted data enters the context via gh run list, gh release view, and git tag output (SKILL.md).
  • Boundary markers: None identified; external data from the repository is interpolated directly into the release drafting process.
  • Capability inventory: The skill has permission to execute git push, gh release edit, and gh workflow run (SKILL.md).
  • Sanitization: There is no explicit sanitization or validation of the data retrieved from repository history or GitHub API before it is presented for release note generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 08:52 PM
Security Audit — agent-trust-hub — run-release