CodeLogician Interfaces

CodeLogician can be used through several interfaces depending on your workflow.

codelogician CLI

The codelogician CLI provides the full CodeLogician platform that is targeted at both developers and agents:

Capabilities include:

• Built-in IML/ImandraX documentation (doc)
• Multi-file reasoning workflows (eval)
• Region decomposition and test generation
• Verification goals

This interface is recommended for:

• coding agents: Claude Code, Codex, Gemini CLI, Cursor Agent with shell access, etc.
• advanced users
• large codebases
• custom workflows
• deep reasoning analysis

→ See: /docs/interfaces/codelogician-cli

VS Code Extension

The VS Code extension integrates CodeLogician into the developer IDE experience.

Features include:

• agent-driven formalization
• interactive reasoning workflows
• seamless IDE integration

→ See: /docs/interfaces/vscode-extension

MCP Server

The MCP server exposes CodeLogician capabilities as tools that can be used by agent frameworks.

This allows:

• AI tools to call CodeLogician reasoning
• integration with agent orchestration systems
• programmatic reasoning workflows

→ See: /docs/interfaces/mcp

Python Library

The Python library provides programmatic access to CodeLogician functionality.

Use this interface when:

• embedding reasoning into applications
• building custom automation
• integrating CodeLogician with internal systems

→ See: /docs/interfaces/python-library

Choosing the Right Interface

As a general rule:

• Use the CLI when working with coding agents, exploring CodeLogician, and full platform capabilities
• Use VS Code for IDE-based development
• Use MCP for agent framework integrations
• Use the Python library for automation and scripting