Skip to content

CLDK CLDK

Codellm-Devkit (aka. CLDK) is a multilingual program analysis framework that bridges the gap between traditional static analysis tools and Large Language Models (LLMs) specialized for code (CodeLLMs). CLDK simplifies multi-language code analysis by providing a unified Python library that integrates outputs from various analysis tools and prepares them for effective use by CodeLLMs.

CLDK streamlines the process of transforming raw code into actionable insights, enabling robust analysis pipelines and seamless integration with tools like WALA, Tree-sitter, LLVM, and CodeQL.

Developer Guide#

Get started with Codellm-Devkit:

  • Quickstart

    Run through an example to quickly set up CLDK and perform multilingual code analysis.

  • Installing cldk

    Install and initialize the cldk Python package to start analyzing your codebases.

  • Python API Reference

    Developer-focused, detailed API reference documentation for cldk.

Why Codellm-Devkit?#

TL;DR

CLDK unifies traditional program analysis tools with CodeLLMs, streamlining multi-language code analysis into a single, cohesive framework.

Current code analysis often involves juggling multiple disjointed tools and workflows. With Codellm-Devkit, you can:

  • Unified: Integrate various analysis tools and CodeLLMs into one cohesive framework.
  • Extensible: Easily add support for new tools and evolving LLM platforms.
  • Streamlined: Simplify the transformation of raw code into structured, actionable insights.

By providing a consistent and extensible interface, CLDK reduces friction and accelerates the development of robust analysis pipelines.

Contact#

For any questions, feedback, or suggestions, please contact the authors:

Name Email
Rahul Krishna [email protected]
Rangeet Pan [email protected]
Saurabh Sihna [email protected]