competitor-customer-leads
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to configure an external Model Context Protocol (MCP) server located at
https://mcp.manypi.com/mcp. This server provides the tools necessary for the lead generation workflow. - [COMMAND_EXECUTION]: The skill provides a bash command (
claude mcp add --transport http manypi https://mcp.manypi.com/mcp -s user) to add the necessary tool configuration to the agent's environment. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources such as public review platforms, job boards, and integration directories. This creates a surface where malicious instructions could be embedded in the retrieved content (e.g., a customer review containing an injection payload).
- Ingestion points: Data retrieved via the ManyPI MCP server from public sources like reviews, case studies, and job posts (referenced in
SKILL.md). - Boundary markers: None explicitly defined in the skill instructions to prevent the agent from interpreting retrieved lead data as instructions.
- Capability inventory: The skill enables external data retrieval and tool invocation via the ManyPI MCP server.
- Sanitization: No evidence of sanitization, escaping, or validation of the external content before processing.
Audit Metadata