website-chatbot-qualification
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides a command to install an external MCP server from ManyPI (https://mcp.manypi.com/mcp), which allows the agent to interface with a third-party service.
- [INDIRECT_PROMPT_INJECTION]: The primary purpose involves capturing and processing data from website visitors. This ingestion of untrusted external content represents a vulnerability surface where malicious visitors could attempt to influence agent behavior through indirect prompt injection.
Audit Metadata