mongodb-mcp-setup

Pass

Audited by Gen Agent Trust Hub on Nov 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands such as env | grep, grep on configuration files, and docker info to inspect the user's environment and determine the current setup status.\n- [PERSISTENCE]: Instructions are provided for users to modify their shell profiles (e.g., ~/.zshrc) to source an environment file, which is a standard method for persisting configuration variables.\n- [INDIRECT_PROMPT_INJECTION]: The skill reads existing configuration from environment variables and local files (~/.codex/config.toml) to assess the current state.\n
  • Ingestion points: Shell environment variables via env and the Codex configuration file.\n
  • Boundary markers: Not present.\n
  • Capability inventory: File reading via grep, shell environment inspection, and local container checks via docker.\n
  • Sanitization: The skill employs sed commands to redact sensitive credential values from the output before the agent processes the information, ensuring secrets do not enter the conversation context.\n- [EXTERNAL_DOWNLOADS]: The skill references and retrieves documentation from the official MongoDB website for installation and troubleshooting purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 17, 2026, 09:10 PM