fix-sentry

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

• Third-party content exposure detected (high risk: 0.70). This skill's workflow (Step 1 and Step 2 in SKILL.md) requires the agent to fetch and read Sentry issue details via sentry_polar and Logfire logs via logfire_polar—external, user-generated/untrusted third-party content—which the agent must interpret to determine root cause and drive fixes/PR actions.