skills/prefecthq/fastmcp/release/Gen Agent Trust Hub

release

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses various command-line tools to manage the repository and release workflow.
  • Evidence: Use of git for branch management and tagging.
  • Evidence: Use of gh (GitHub CLI) for release creation, PR management, and API queries.
  • Evidence: Use of uv to execute local release scripts.
  • [EXTERNAL_DOWNLOADS]: The skill fetches external tools and verifies deployment status via the network.
  • Evidence: npx --yes mint@latest downloads the latest Mintlify CLI for documentation validation.
  • Evidence: curl calls to pypi.org and the project's official domain to verify successful publication.
  • [INDIRECT_PROMPT_INJECTION]: The release script processes external PR titles which could contain malicious instructions or breaking characters.
  • Ingestion points: scripts/changelog_entry.py fetches release notes containing PR titles from the GitHub API.
  • Boundary markers: The content is structured within MDX <Update> tags.
  • Capability inventory: The script writes generated content to local .mdx files, which are later committed and pushed to the repository.
  • Sanitization: The script implements an escape_mdx function designed to wrap characters that could be interpreted as JSX (like < followed by a digit, or {/}) in backticks, mitigating potential documentation injection or build failures.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 03:59 AM