self-improvement

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

• Third-party content exposure detected (high risk: 0.90). The "Learning from GitHub PRs and Issues" section explicitly instructs the agent to run gh CLI commands to fetch public PRs, reviews, and comments (user-generated content) and to extract lessons/promote them to CLAUDE.md, so untrusted third‑party content is read and can materially influence the agent's actions.