add-context-gateway
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the discovery of existing MCP configurations by analyzing project files such as
.mcp.jsonand external tool schemas. While this ingestion represents a surface for indirect prompt injection, it is a core requirement for the gateway setup process. Ingestion points: Project configuration files (.mcp.json,.cursor/mcp.json) and schemas from external MCP servers. Boundary markers: Not explicitly defined in instructions. Capability inventory: Shell command execution via theirisCLI and file writing to.envand other configuration files. Sanitization: Not explicitly mentioned, as the data is used for standard configuration tasks. - [EXTERNAL_DOWNLOADS]: The skill directs the installation of the
redis-irispackage. This is an official tool from the vendor 'redis', consistent with the skill's stated purpose. - [COMMAND_EXECUTION]: The instructions involve multiple CLI operations using the
iristool for provisioning and verifying services. These commands are structured to prioritize security, utilizing environment variables for credentials and file redirection to prevent accidental exposure of secrets in the output transcript.
Audit Metadata