global-agriculture-food-security-data

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

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow for query_agriculture_data fetches agriculture/food-security indicator data from external third-party sources (World Bank WDI and FAO), and the tool’s returned JSON is ingested into the agent’s LLM context as readable indicator text—an outsider-authored data source path.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). This skill explicitly directs runtime calls to AgentPMT's MCP to fetch live schema/instructions (e.g., via agentpmt-tool-search-and-execution against https://api.agentpmt.com/mcp/ using get_schema/get_instructions), and those returned instructions are treated as authoritative and would directly control agent prompts.