didit-mcp-server
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads the official vendor package
@didit-protocol/mcp-serverand themcp-remoteutility from the npm registry usingnpxduring installation and setup. - [REMOTE_CODE_EXECUTION]: Executes remote code from the downloaded npm packages to provide the MCP server functionality and connect the AI agent to the Didit platform.
- [COMMAND_EXECUTION]: Utilizes shell commands (e.g.,
npx,claude mcp add) to register and run the server in various client environments. - [INDIRECT_PROMPT_INJECTION]: The skill ingests JSON data from the Didit platform APIs, creating a potential surface for indirect prompt injection. This risk is managed through explicit human-in-the-loop requirements; the agent must seek user approval and pass a
confirm: trueparameter for all destructive actions, such asdidit_session_deleteordidit_org_reveal_application_api_key.
Audit Metadata