Pick up new bugs automatically under team rules
Research prototype for reference only, not an official DeepSeek product. Task: Research prototype for reference only, not an official DeepSeek product. This code grew out of our team's day-to-day development work, with much of the engineering done by coding agents. Treat it as research code, not production-ready software. It may still contain bugs, and its coding style and engineering practices may be inconsistent. Review and adapt it before using it in your own environment. We have adapted it for publication, so it is not identical to our internal version or the implementation described in Exploring Super-agent Design: Towards a More Autonomous Agent System. Some designs discussed in the article are directions for future development and are not implemented in this prototype. dsh-superagent is a proactive coding agent for one GitHub repository, packaged as a bundle for DeepSeek Harness. It listens to GitHub and, optionally, Feishu. When someone allowed asks the agent for work, it opens a work item, works on it in its own agent session and directory, and answers in the original conversation, for example with a pull request. A web panel shows what needs a person. It is also a way of working with an agent: take as little human attention as possible, and give as much value as possible. An agent that can act for a whole team can also waste a whole team's attention, so it acts only when asked or when a rule the team wrote allows it, and speaks only when someone must act or can use a result. docs/000-philosophy.md explains the principle and what it asks of the operator. Setup Setup has many steps, so an agent does it with you. Ask your coding agent: Read SETUP.md in this repository and set up dsh-superagent with me. It asks one short question at a time, shows the full plan once, and then installs and configures everything. SETUP.md is what it follows. Documentation Read the numbered documents in order; each builds only on the ones before it. 000 Philosophy 001 Domain model 002 Authorization 003 Execution 004 Prompts 005 Panel 006 Deployment: manual setup, configuration, upgrades and troubleshooting 007 Architecture 008 Cookbook: adding a Source Development pnpm check runs the typecheck, the tests and the build. License MIT; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: The source is a research prototype, not an official product, and differs from internal versions or described future designs. Setup is complex and agent-assisted via SETUP.md. The agent requires specific authorization rules and integration with GitHub and optionally Feishu. Human review is mandatory before merging changes, as the agent is designed to minimize attention waste but cannot fully replace human oversight..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
Research prototype for reference only, not an official DeepSeek product.
Workflow
Research prototype for reference only, not an official DeepSeek product. This code grew out of our team's day-to-day development work, with much of the engineering done by coding agents. Treat it as research code, not production-ready software. It may still contain bugs, and its coding style and engineering practices may be inconsistent. Review and adapt it before using it in your own environment. We have adapted it for publication, so it is not identical to our internal version or the implementation described in Exploring Super-agent Design: Towards a More Autonomous Agent System. Some designs discussed in the article are directions for future development and are not implemented in this prototype. dsh-superagent is a proactive coding agent for one GitHub repository, packaged as a bundle for DeepSeek Harness. It listens to GitHub and, optionally, Feishu. When someone allowed asks the agent for work, it opens a work item, works on it in its own agent session and directory, and answers in the original conversation, for example with a pull request. A web panel shows what needs a person. It is also a way of working with an agent: take as little human attention as possible, and give as much value as possible. An agent that can act for a whole team can also waste a whole team's attention, so it acts only when asked or when a rule the team wrote allows it, and speaks only when someone must act or can use a result. docs/000-philosophy.md explains the principle and what it asks of the operator. Setup Setup has many steps, so an agent does it with you. Ask your coding agent: Read SETUP.md in this repository and set up dsh-superagent with me. It asks one short question at a time, shows the full plan once, and then installs and configures everything. SETUP.md is what it follows. Documentation Read the numbered documents in order; each builds only on the ones before it. 000 Philosophy 001 Domain model 002 Authorization 003 Execution 004 Prompts 005 Panel 006 Deployment: manual setup, configuration, upgrades and troubleshooting 007 Architecture 008 Cookbook: adding a Source Development pnpm check runs the typecheck, the tests and the build. License MIT
Reported result
No measured outcomes or success metrics are reported in the source. The prototype is explicitly labeled as research code, not production-ready, and may contain bugs or inconsistent engineering practices. Acceptance check: Verify that the agent correctly opens a work item and submits a pull request in response to an authorized GitHub issue, with the web panel accurately flagging any steps requiring human approval.
Human role
The original post does not say how a person stayed involved.
Limitations
The source is a research prototype, not an official product, and differs from internal versions or described future designs. Setup is complex and agent-assisted via SETUP.md. The agent requires specific authorization rules and integration with GitHub and optionally Feishu. Human review is mandatory before merging changes, as the agent is designed to minimize attention waste but cannot fully replace human oversight.
Try it
I want to set up the dsh-superagent prototype for my team's GitHub repository [replaceable repo URL]. First, read SETUP.md and guide me through the installation step-by-step, asking one question at a time. Configure the agent to listen to GitHub issues and optionally Feishu [replaceable channel]. Define a rule where the agent only acts when a user with role [replaceable role] comments '/agent fix'. Implement the workflow: when triggered, the agent creates a work item, processes it in an isolated directory, and submits a pull request. Ensure the web panel displays any actions requiring human approval. Before running any real tasks, perform a dry run on a test issue [replaceable test issue number] and show me the proposed pull request and panel status for my confirmation. Do not merge or publish anything without my explicit approval. If setup fails, report the specific error from SETUP.md and ask for clarification on missing dependencies.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.