marketing-agency-leads
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides a command to install a Model Context Protocol (MCP) server:
claude mcp add --transport http manypi https://mcp.manypi.com/mcp -s user. This installation source is hosted on the domain associated with the skill's author (manypicom) and represents standard functionality for the vendor's toolset. - [INDIRECT_PROMPT_INJECTION]: The skill's primary function requires the agent to browse and extract data from third-party websites such as agency directories, team pages, and job boards.
- Ingestion points: Untrusted data enters the agent's context from external URLs visited during the lead qualification process (SKILL.md).
- Boundary markers: There are no explicit instructions or delimiters provided to prevent the agent from following instructions embedded within external web content.
- Capability inventory: The agent utilizes the
manypitool to perform search and web data extraction. - Sanitization: No validation or sanitization of the retrieved external content is mentioned.
- This activity creates a potential surface for indirect prompt injection if a visited website contains hidden instructions designed to influence the agent's behavior.
Audit Metadata