Home of CodeT5: Open Code LLMs for Code Understanding and Generation

CodeTF: One-stop Transformer Library for State-of-the-art Code LLM

Official code of our work, Unified Pre-training for Program Understanding and Generation [NAACL 2021].

This repository provide the studies on the security of language models for code (CodeLMs).

Source code for EMSE 2023 paper "Zero-Shot Code Representation Learning via Prompt Tuning"

Conversational AI agent for repository analysis using Git history patterns. Reveals code quality evolution, complexity trends, and decay forecasting through natural language interactions. Next-generation temporal intelligence platform.

A code analysis tool that generates dependency graphs from multi-language codebases, enabling AI agents to explore and understand code structure through MCP integration. Features real-time file watching, dynamic instrumentation, and security scanning.

🌈 Paper-implementations in Code Search (Baseline).

An AI-first code analysis and safe refactoring platform with a Python SDK and JSON API. Designed for AI agents with a zero-index architecture for 100x speed and unbreakable safety.

CodebaseIQ Pro: The #1 MCP server tool for large codebase analysis; saving you time, money, & headaches. 🧠 With a single command, give your coding agent complete codebase knowledge in one shot, preventing costly errors from incomplete understanding and redundant training per fresh conversation.

🚀 Intelligent MCP server that gives AI assistants persistent memory, advanced code analysis, and context-aware development tools

Cursor AI SSOT & Rule Engine Template: A comprehensive framework for integrating Cursor AI's advanced features into any project. Includes Single Source of Truth (SSOT) tracking, custom project rules, and Model Context Protocol (MCP) with Knowledge Graph for intelligent context management.