release-status

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses a Node.js helper to run git and gh commands for gathering repository information. All command arguments derived from user input are validated, and the tool uses non-shell execution for security.
  • [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection by processing external data from commit history and GitHub PRs.
  • Ingestion points: Untrusted commit subjects, bodies, and PR titles are ingested from the repository environment via scripts/release-status.mjs.
  • Boundary markers: No explicit delimiters are used in the generated report to distinguish untrusted content from diagnostic instructions.
  • Capability inventory: The skill can execute shell commands (git, gh) and read local repository files.
  • Sanitization: Input processing includes regex-based classification of commit messages, which restricts the logic to recognizing Conventional Commit patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 12:05 PM
Security Audit — agent-trust-hub — release-status