feed · watching · stats · help with your tokens
deepseek-ai/dsh-superagent-prototype
dsh-superagent is a research prototype for a proactive coding agent that works on a single GitHub repository using DeepSeek Harness.
It aims to maximize value while minimizing human attention by acting only when asked or permitted by team rules, serving developers who want autonomous coding assistance without constant oversight. It listens to GitHub and Feishu, opens work items, executes tasks in isolated agent sessions, and responds with pull requests or conversation updates.
Similar
- JustVugg/colibri
colibri is a C inference engine that runs very large open mixture-of-experts models on the machine you already have.
- anthropics/FlashProve
FlashProve is a research prototype for zero-knowledge proofs of AI inference that allows providers to prove model identity without revealing weights.
- NVIDIA/auto-ontology
An open context layer that turns database metadata into an ontology, so AI agents answer questions about enterprise data with fewer wrong answers.
- pgEdge/ai-dba-workbench
A PostgreSQL monitoring system that combines traditional and AI-based monitoring for instances v14+ including Supabase and Amazon RDS.
- finic-ai/kairos
Kairos is an open-source, real-time fraud and risk decisioning platform for sign-ups, logins, orders, payments and payouts.
- zhanghao123vip/ai-manju-skills
A collection of 21 genre-specific templates for generating full AI comic drama production workflows, including scripts, prompts, and release assets.
- mycelium-hq/ai-brain-starter
AI Brain Starter is a verification harness for Claude Code that provides persistent memory, deterministic guards, and session lifecycle management to prevent context corruption.
- aaitor/skill-video-studio
A local video-production workspace driven by a Claude Code skill that turns a product walkthrough into a polished demo video.
sources
Claims below include README evidence. A matching quote or worker agreement does not establish that a claim is correct. Category majority among 2 workers. Version 1.
dsh-superagent is a research prototype for a proactive coding agent that works on a single GitHub repository using DeepSeek Harness.
dsh-superagent is a proactive coding agent for one GitHub repository, packaged as a bundle for DeepSeek Harness.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
It aims to maximize value while minimizing human attention by acting only when asked or permitted by team rules, serving developers who want autonomous coding assistance without constant oversight.
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
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
It listens to GitHub and Feishu, opens work items, executes tasks in isolated agent sessions, and responds with pull requests or conversation updates.
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.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Proactive coding agent for a single GitHub repository packaged for DeepSeek Harness.
dsh-superagent is a proactive coding agent for one GitHub repository, packaged as a bundle for DeepSeek Harness.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Includes a web panel to display tasks requiring human attention.
A web panel shows what needs a person.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Operates only when explicitly asked or permitted by team rules, minimizing unnecessary interruptions.
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.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Setup is guided by an agent following SETUP.md, asking one question at a time.
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.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Development workflow uses pnpm check for typechecking, testing, and building.
pnpm check runs the typecheck, the tests and the build.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Monitors GitHub repositories and optionally Feishu for incoming work requests or triggers.
It listens to GitHub and, optionally, Feishu.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Executes work items autonomously in isolated sessions and directories, responding to users via pull requests or the original conversation channel.
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.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Provides a web dashboard to visualize tasks requiring human attention or intervention.
A web panel shows what needs a person.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Designed to minimize human attention while maximizing value by acting only when explicitly asked or when team-defined rules permit.
take as little human attention as possible, and give as much value as possible.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Operates under strict constraints to act only upon request or team rules, and communicates only when action or result usage is required.
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.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Prevents team attention waste by restricting autonomous actions to explicit requests or pre-defined team rules.
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.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Research prototype only, not production-ready; may contain bugs and inconsistent engineering practices.
This code grew out of our team's day-to-day development work... Treat it as research code, not production-ready software. It may still contain bugs, and its coding style and engineering practices may be inconsistent.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Not identical to the internal version or the full implementation described in the associated research article.
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*.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Some designs from the research article are not implemented in this prototype.
Some designs discussed in the article are directions for future development and are not implemented in this prototype.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: This is a research prototype, not an official product, and may contain bugs or inconsistent engineering practices.
Research prototype for reference only, not an official DeepSeek product.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: The codebase may include bugs and inconsistent coding styles, requiring review and adaptation before production use.
It may still contain bugs, and its coding style and engineering practices may be inconsistent.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Certain designs mentioned in the associated article are future directions and are not currently implemented in this prototype.
Some designs discussed in the article are directions for future development and are not implemented in this prototype.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Users must review and adapt the code before deploying it in their own environments due to its prototype nature.
Review and adapt it before using it in your own environment.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
ai-ml