glab

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly calls and ingests content from public GitLab endpoints (e.g., references/api-and-advanced.md and the "Common Patterns for Agents" in SKILL.md that use glab api, glab issue list -F json, glab mr list and glab snippet view), which are user-generated/untrusted (issues, MRs, comments, snippets) and the docs show workflows that parse that content and then perform follow-up actions (closing issues, calling APIs, triggering pipelines), so third‑party content can materially influence agent actions.