pp-shelters

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

• Third-party content exposure detected (high risk: 0.85). The required workflow fetches the live FEMA OpenShelters feed at runtime and returns it as JSON ({source, fetched_at, data}), so the CLI’s fetched shelter records (outsider-authored public web content) can be ingested into the agent’s LLM context via the --agent/stdout output path.