prs

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 is PR/queue triage and merge-conflict handling, which necessarily ingests outsider-authored PR/issue text (e.g., PR description, comments, and diff content from other authors) into the agent’s LLM context via the PR/queue inputs it processes.