release

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to manage the software lifecycle.
  • Runs node -p to extract version information from local package.json files in references/tools/changesets.md.
  • Calls npm publish, git push, and gh release to synchronize code and artifacts with remote registries and repositories as defined in the release standards.
  • Employs tar to verify the integrity and contents of built packages before publication.
  • [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection by processing external text data.
  • Ingestion points: Reads and summarizes CHANGELOG.md, git log, and PR descriptions to generate release notes in references/release-content.md.
  • Boundary markers: No explicit instruction for the agent to use delimiters or safety prompts when handling ingested text.
  • Capability inventory: Has the authority to publish to npm and push to GitHub.
  • Sanitization: Lacks specific content sanitization for external strings.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with external registries and services to manage project tooling.
  • Downloads and installs packages like @changesets/cli, publint, and @arethetypeswrong/cli from the official npm registry.
  • Fetches configuration schemas from the unpkg.com CDN.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 07:33 AM