track-climate-hazards

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes a literal API key value and instructs callers to put an API key verbatim in the X-WorldMonitor-Key header, which encourages the LLM to handle or emit secret values directly.

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md describes runtime calls to external WorldMonitor API endpoints (e.g., GET /api/climate/v1/list-climate-disasters|anomalies|news) whose response fields (event names, headlines, source text) are explicitly sourced from external feeds and may contain untrusted free text that would be ingested into the agent’s LLM context for analysis.