Omni

Exposes Omni’s 55 scanners as AI-callable tools for end-to-end codebase observation and governance.

other

GitHub Stars

other

Language

3 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "pantheon-ladderworks-omni-scanner": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/inspector",
        "mcp-server-omni-governance"
      ],
      "env": {
        "OMNI_ROOT": "\"/workspace/project\"",
        "OMNI_WORKSPACES": "\"/workspace/one,/workspace/two\"",
        "OMNI_REPO_INVENTORY": "\"/workspace/inventory.json\""
      }
    }
  }
}

Omni MCP Server exposes Omni’s 55 scanners as AI-callable tools, enabling you to query, orchestrate, and inspect your codebase through an MCP-compatible assistant. It provides ready-made, plug-and-play access to comprehensive codebase observation capabilities without modifying files or disrupting builds.

How to use

You interact with the MCP server through an MCP client or AI assistant that can call Omni’s scanners as endpoints. Start a local MCP server instance to expose Omni’s observational tools, then instruct your client to request scans, introspection, or reports. You can run scans against project directories, generate maps of dependencies, and retrieve governance-oriented insights through structured tool invocations.

How to install

Prerequisites: ensure you have Python 3.8+ and a working Python environment. If you plan to use the Node-based MCP client approach, also have Node.js installed.

# Option 1: Use MCP server via Node.js client (npx)\n# The absolute easiest way to use the MCP Server is via npx (requires Node.js):\nnpx -y @modelcontextprotocol/inspector mcp-server-omni-governance

# Option 2: Run the MCP server from source (Python)\n# The MCP server auto-discovers all registered scanners\npython -m mcp_server.omni_mcp_server

Configuration and starting the server

Two primary ways to run the MCP server are supported: a Node.js-based client invocation and a Python-based local server. Choose the method that best fits your development workflow. No external backend is required for standalone usage.

Security and access

Keep your MCP endpoints private to your development environment or secure them with appropriate access controls if you expose them over a network. Do not run with elevated privileges unless necessary, and monitor any environment that provides access to scanning capabilities to prevent unauthorized observations of codebases.

Troubleshooting

If you encounter issues starting the MCP server, ensure Python is available on your PATH for the Python invocation, or verify Node.js is installed for the npx invocation. Check that your environment has network access if you are using remote tooling, and confirm that the MCP server process has the necessary permissions to load scanners.

Notes

The MCP server exposes a suite of scanners as AI-callable tools. You can request scans, introspection, mapping, and reporting through your MCP client, enabling practical governance-oriented observations of your codebase without altering any source files.

Built by

VeilStrat

AI signals for GTM teams