draft-release-notes

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow uses gh pr list / gh api to fetch PR metadata (titles, authors, labels) from GitHub for the release range, and those PR titles/metadata are outsider-authored text that the agent then ingests into the LLM context to write the release notes.