pr-workflow

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and ingests user-generated review/comments from public GitHub via the gh API and miner.py (see references/feedback.md and references/miner.md which call gh api and scripts/miner.py to retrieve reviews, inline comments, and issue comments) and then reads and acts on that content (validate/fix code, generate rules), exposing the agent to untrusted third-party input that could contain indirect prompt-injection instructions.