newsletter

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and ingests public GitHub PR content (e.g., "Matches GitHub URL or #\d+ -> PR" and instructions to use gh pr view / gh pr diff to read PR descriptions, diffs, and review comments), which are user-generated, untrusted third-party sources that the agent must read and use to decide newsletter content.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches PR/diff and repo content at runtime (via commands like "gh pr view", "gh pr diff", "git diff" or by matching GitHub PR URLs such as https://github.com///pull/), and injects that external content into the model context to drive newsletter generation, so external content can directly control prompts.