social-pulse-monitor

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text is ingested via the runtime keyless connectors (e.g., tavily.py, gdelt.py, bluesky.py, discourse.py, hn.py) that fetch public web/forum/news content and then feed the fetched mention text into the agent’s LLM context for triage and spike attribution.