helix-mcp-icp-intelligence
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references an external npm package @shashwatgtmalpha/icp-intelligence-mcp and a GitHub repository at https://github.com/shashwatgtm/icp-intelligence-mcp. Both resources are maintained by the author shashwatgtm.
- [COMMAND_EXECUTION]: Includes a CLI command for users to register the remote MCP server: 'claude mcp add --transport http icp-intelligence https://icp-intelligence.gtmhelix.com/mcp'.
- [DATA_EXFILTRATION]: The skill transmits data to a hosted endpoint at https://icp-intelligence.gtmhelix.com/mcp for processing. Data including customer lists and interview notes is sent to this external service as part of the tool's core functionality.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by ingesting untrusted data. 1. Ingestion points: 'icp_deep_dive' (customer lists) and 'icp_interview_synthesizer' (raw interview notes) in SKILL.md. 2. Boundary markers: No explicit delimiters or instructions to ignore embedded content are provided. 3. Capability inventory: The skill utilizes 'web_search', 'web_fetch', and a remote HTTP MCP server. 4. Sanitization: No sanitization or validation of external data is specified before processing.
Audit Metadata