dsh-superagent vs SWE-agent
Which should you use? A side-by-side look at what each one does, who it suits, what it costs, how hard it is to set up, and where each falls short.
Open source · MIT
A research prototype of a coding agent that watches one GitHub repo and acts when asked.
Open source · MIT
An AI agent that tries to fix GitHub issues automatically, using the language model you choose.
The short answer
Choose dsh-superagent if you need…
- Engineering teams exploring autonomous coding agents on a single GitHub repository
- Developers studying how agent work items, authorization and a review panel can fit together
- Researchers reading the accompanying design article alongside working code
- Operators who want an agent to act only when a team rule or a person allows it
Choose SWE-agent if you need…
- Researchers studying autonomous coding agents or running SWE-bench experiments
- Developers who want to experiment with an LLM attempting issue fixes in their own repositories
- Students learning how agent-computer interfaces work, since the README describes it as simple and hackable
- People who want a configurable agent controlled by a single YAML file
Side-by-side
| dsh-superagent | SWE-agent | |
|---|---|---|
| Type | AI model / agent | AI model / agent |
| Price | Free and open source | Free and open source |
| Licence | MIT | MIT |
| GitHub stars | 11k | 20.7k |
| Setup difficulty | Advanced | Intermediate |
| How you get it | Clone and build from source | Clone and build from source |
| Main language | TypeScript | Python |
| Last code update | 6 days ago | 10 days ago |
What each one is
dsh-superagent
dsh-superagent is a proactive coding agent built for one GitHub repository, packaged as a bundle for DeepSeek Harness. It listens to GitHub and optionally Feishu; when an allowed person asks for work, it opens a work item, works on it in its own session and directory, and replies in the original conversation, for example with a pull request. A web panel shows what needs a person's attention. The README describes it as a research prototype, not an official DeepSeek product, and not production-ready.
SWE-agent
SWE-agent takes a GitHub issue and lets a language model you pick (the README mentions GPT-4o and Claude Sonnet 4 as examples) use tools to try to fix it in a real repository. It is also described as usable for offensive cybersecurity (capture-the-flag) challenges and custom coding tasks. It is a research project from Princeton and Stanford, and its README says its main development has moved to the simpler mini-SWE-agent.
Best for
dsh-superagent
- Engineering teams exploring autonomous coding agents on a single GitHub repository
- Developers studying how agent work items, authorization and a review panel can fit together
- Researchers reading the accompanying design article alongside working code
- Operators who want an agent to act only when a team rule or a person allows it
SWE-agent
- Researchers studying autonomous coding agents or running SWE-bench experiments
- Developers who want to experiment with an LLM attempting issue fixes in their own repositories
- Students learning how agent-computer interfaces work, since the README describes it as simple and hackable
- People who want a configurable agent controlled by a single YAML file
Strengths
dsh-superagent
- Clear stated philosophy of minimizing human attention, documented in docs/000-philosophy.md
- Ordered documentation set covering domain model, authorization, execution, prompts and architecture
- Uses the MIT license, which permits reuse with few conditions
SWE-agent
- The README reports state-of-the-art results on SWE-bench among open-source projects
- Configuration is handled by a single YAML file and the README says it is fully documented
- MIT licensed with a public docs site, Slack, and a Codespaces option to try it in a browser
Not ideal for
dsh-superagent
- Teams needing production-ready software, since the README warns it may contain bugs and inconsistent practices
- Anyone wanting a simple install in minutes, since the README says setup has many steps and recommends an agent-guided setup
- Users who need features described in the article, since the README says some designs are not implemented in this prototype
SWE-agent
- Teams wanting the project's actively recommended tool: the README says most development has moved to mini-SWE-agent, which it recommends instead
- Anyone needing a polished, finished product: the project is framed as research code
- People who cannot pay for or access a language model API, since the README assumes you bring your own model
Trade-offs
dsh-superagent
- The README explicitly labels it research code that should be reviewed and adapted before use
- The README says it is not identical to the internal version or the article's implementation
- Setup has many steps and the full install details are not included in the truncated README
SWE-agent
- The README says mini-SWE-agent has superseded it and is recommended instead going forward
- The EnIGMA cybersecurity mode currently points users to older version 0.7 rather than 1.0
- Results depend on the language model and API you supply; the README does not state costs or hardware needs
Getting started
dsh-superagent
Clone the repository
git clone https://github.com/deepseek-ai/dsh-superagent-prototypeRun the project checks (typecheck, tests, build)
pnpm checkSWE-agent
Clone the repository
git clone https://github.com/SWE-agent/SWE-agent.gitFrequently asked questions
What is the difference between dsh-superagent and SWE-agent?
dsh-superagent: A research prototype of a coding agent that watches one GitHub repo and acts when asked. SWE-agent: An AI agent that tries to fix GitHub issues automatically, using the language model you choose. dsh-superagent suits engineering teams exploring autonomous coding agents on a single github repository, while SWE-agent suits researchers studying autonomous coding agents or running swe-bench experiments.
Are dsh-superagent and SWE-agent free?
dsh-superagent: Free and open source. SWE-agent: Free and open source.
Which is easier to set up, dsh-superagent or SWE-agent?
SWE-agent is easier to get started with (rated intermediate); dsh-superagent is rated advanced and usually needs more setup.
Is dsh-superagent or SWE-agent more popular?
SWE-agent has more GitHub stars (20.7k) than dsh-superagent (11k). Popularity isn't the same as fit — check which one matches your use case above.
What are other alternatives to dsh-superagent and SWE-agent?
Other options people compare them with include OpenHands, DeerFlow, Open SWE, RD-Agent, Ponytail.
Keep comparing
Built from our catalogue entries for both products: open-source details come from each project's README (AI-assisted, refreshed weekly) and GitHub data; AI-tool details are editorial. Spotted something wrong? Let us know from either product's page.