fix-false-positive

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly fetches and parses public, user-generated GitHub issue bodies via gh api (see "Step 1" and "2.1" where it calls gh api repos/{OWNER}/{REPO}/issues and gh api .../issues/{N}) and then uses that content to drive rule edits, validation, commits, and PR creation, so untrusted issue text can materially influence agent actions.