openkrak-mcp

MCP server for developer tooling. Performs static analysis on a software repository via the Dorchester engine and delivers a structured context brief to the language model before any file is accessed.

Supported: OpenCode (TUI + Desktop). Claude Code, Codex, and Cursor support in development.

How It Works

OpenKrak runs the Dorchester engine — a six-step static analysis pipeline — on your repository before the language model executes any task.

Repository
    │
    ▼
DeepStrike          — Full static analysis: every import, export, coupling resolved
    │
    ▼
Hotspot Registry    — Files ranked by complexity, change frequency, cognitive load
    │
    ▼
Correlation Engine  — Dependency chains traced across the entire codebase
    │
    ▼
Blast Radius        — Ripple effects of any change mapped before commit
    │
    ▼
Execution Gate      — Analysis validated and normalized
    │
    ▼
Mahadata            — Structured brief delivered to the language model

All analysis runs locally. Your codebase does not leave your machine.

Related MCP server: SGraph MCP Server

Install

npm install -g openkrak-mcp

Requires Node.js ≥ 18.

Setup

OpenKrak works with both OpenCode TUI and OpenCode Desktop. The config file and setup steps are identical for both.

Config file location

Step 1 — Add OpenKrak to your config

Open the config file and add the mcp block:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "openkrak": {
      "type": "local",
      "enabled": true,
      "command": ["npx", "openkrak-mcp"]
    }
  }
}

Step 2 — Restart OpenCode

Close and reopen OpenCode (TUI or Desktop). Confirm openkrak Connected appears in the MCP panel.

OpenCode Desktop: MCP status is shown in the bottom status bar. Green dot = connected.
OpenCode TUI: MCP status is shown in the top-right panel.

Step 3 — Run your first analysis

Open any repository in OpenCode and send a message:

analyze_repo on /absolute/path/to/your/repo

Or just start a coding task normally — OpenKrak runs automatically before the model reads any file.

Tools

Example prompts

analyze_repo on /path/to/repo with objective "refactor the auth module"

Run get_hotspots on /path/to/repo — which files should I avoid touching?

blast_radius on /path/to/repo for file src/api/routes/user.ts

Enable / Disable

Set "enabled" in your config and restart OpenCode:

"enabled": true   // on
"enabled": false  // off

Pro License

Free tier is active by default — no account required.

To activate Pro, add your key to the config:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "openkrak": {
      "type": "local",
      "enabled": true,
      "command": ["npx", "openkrak-mcp"],
      "env": {
        "OPENKRAK_KEY": "your-license-key"
      }
    }
  }
}

License keys: openkrak-web.vercel.app. Delivered via email after purchase. No subscription, no auto-renewal.

Notes

• Static analysis only. No AI inference in the pipeline.

• Anonymous usage telemetry is collected (query count, tool name, error events). No source code or file contents are transmitted.

• License validation requires a network call to the license server on each tool invocation.

Repository

github.com/FrnzJulianBergmann/openkrak

MIT License — © 2026 Faiz Hamizan / Challanger Absolute Advance