SWE-agent SWE-agent
SWE-agent is a specialized autonomous agent designed to read GitHub issues, navigate repositories, and generate tested pull requests. It utilizes a purpose-built Agent-Computer Interface (ACI) to interact with codebases, effectively turning an LLM into an autonomous software engineer for bug fixing and maintenance tasks.
The project is currently in maintenance-only mode, as the development team has shifted all primary effort to mini-swe-agent, a simpler, high-performance successor.
As of
- Agent-Computer Interface (ACI) for efficient codebase navigation
- Autonomous bug fixing with automatic PR generation
- Docker-based sandbox for safe, reproducible execution
- Highly configurable behavior via a single YAML file
- EnIGMA mode for offensive cybersecurity and CTF challenges
Researchers and developers seeking a transparent, hackable, and state-of-the-art autonomous tool for fixing well-defined bugs and performing code maintenance.
Users who prefer a polished, managed, or IDE-integrated experience, or those looking for general-purpose features like scaffolded feature development.
The software is free and open-source under the MIT license. Users pay only for their own LLM API tokens (e.g., OpenAI or Anthropic) and compute/infrastructure costs.
While it remains a landmark research project and a top-performing open-source agent, users are officially recommended to use mini-swe-agent for new projects due to its simplicity and active development status.
Is SWE-agent free?
Yes - SWE-agent is Open Source. The software is free and open-source under the MIT license. Users pay only for their own LLM API tokens (e.g., OpenAI or Anthropic) and compute/infrastructure costs.
Is SWE-agent open source?
Yes - SWE-agent is open source. The software is free and open-source under the MIT license. Users pay only for their own LLM API tokens (e.g., OpenAI or Anthropic) and compute/infrastructure costs.
Who is SWE-agent best for?
Researchers and developers seeking a transparent, hackable, and state-of-the-art autonomous tool for fixing well-defined bugs and performing code maintenance.
Who is SWE-agent not ideal for?
Users who prefer a polished, managed, or IDE-integrated experience, or those looking for general-purpose features like scaffolded feature development.
What are the best SWE-agent alternatives?
The closest SWE-agent alternatives on ai.dosa.dev are Jules, OpenHands, Sweep - all listed under Autonomous & Async Agents.
Who makes SWE-agent?
SWE-agent is developed by Princeton NLP (in collaboration with researchers from Stanford University). It is listed in the Autonomous & Async Agents category on ai.dosa.dev.
Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.
Content on this page is AI-generated. Please verify details with the vendor's website for accuracy.