Open-Source AI Coding Agents
AI coding agents support software development by analyzing source code, editing files, running commands, and testing changes.
Typical tasks include implementing features, fixing bugs, refactoring, and writing tests. Coding agents carry out these development tasks themselves. Depending on the product, you can use them in the terminal, in an IDE, on the desktop, or in the browser.
Related Categories
Tools for AI agent management mainly coordinate existing agents and their tasks rather than carrying out development tasks themselves. For building your own agents and workflows, there are AI agent platforms.
Personal AI assistants handle personal and recurring tasks with configured tools and approvals. AI chat interfaces, by contrast, focus on conversations, writing, and research.
Features
Below is an overview of common features of AI coding agents.
Analyze source code and project structure
Implement changes across multiple files
Run commands, tests, and builds
Investigate and fix bugs
Review and explain code changes
Use project instructions and reusable skills
Connect external tools via MCP
Control permissions and approvals for actions
Products
Below is an overview of popular open-source software in the category «Open-Source AI Coding Agents». The list is sorted alphabetically.
Well-known proprietary examples include ChatGPT Desktop App (Codex) and Claude Code.