Code quality tools have an infrastructure problem. SonarQube needs a server. CodeClimate needs a cloud connection. Even lightweight linters need CI pipeline configuration.
Truemend takes a fundamentally different approach: it runs as an MCP server over stdio, directly inside your development environment.
What is MCP?
Model Context Protocol (MCP) is an open standard for connecting AI assistants to external tools. When your IDE’s AI assistant (Claude Code, Cursor, Copilot) needs to analyze code quality, it calls Truemend through MCP: no API key and no cloud dependency for the analysis itself.
Why This Matters
No server round-trip. Analysis happens locally, on your machine.
Minimal configuration. Point Truemend at a directory. That’s it. No YAML files, no project setup, no onboarding wizard.
Analysis stays local. Code analysis runs on your machine and sends your source nowhere. Licensing, account services and definition updates are separate network activity. What your editor's AI assistant does with the results it receives depends on that client and its configuration.
Works offline. On a plane? In a secure facility? Truemend doesn’t care. It’s a local Python package: install once, and analysis runs without a network connection.
How We Use It in Engagements
During a consulting engagement, we configure Truemend as an MCP server in your team’s IDE. Every developer gets real-time code quality feedback as they work: not after they push, not in a CI pipeline, but right there in the editor.
This means quality improvements compound: developers learn the patterns, the codebase gets cleaner with every commit, and our audit findings shrink sprint over sprint.
The One-Command Advantage
Truemend ships as a pip-installable Python package: pip install truemend. No Docker container. No separate server. No account to create.
That’s the entire tool. Anti-patterns, detection engine, fix engine, reporting: all in one package, one command away.