repowiserepowise

deepseek-ai / dsh-superagent-prototype Architecture and code health

Towards a More Autonomous Agent System

★ 22,260stars on GitHubTypeScriptMIT

Repowise maps public repos from their source. Not affiliated with deepseek-ai. What is Repowise?

github.com/deepseek-ai/dsh-superagent-prototypeIndexed at 8c32a9d

What is dsh-superagent-prototype?

dsh-superagent-prototype is a research prototype of a proactive coding agent for one GitHub repository, packaged as a bundle for DeepSeek Harness. It listens to GitHub and optionally Feishu for requests from people who are allowed to ask for work. For each request it opens a work item, runs it in its own agent session and directory, and replies in the original conversation, for example with a pull request. The core is written in TypeScript and packaged with npm.

Written by a model from the README and the source, like the other summaries and answers on this page

206 files and 1,784 symbols in 4 modules, led by TypeScript (171 files) and Python (1).

Explore dsh-superagent-prototype

Ask about this code in your agent

This connects Repowise to dsh-superagent-prototype's code. It does not install dsh-superagent-prototype.

Useful if you want to change or fork dsh-superagent-prototype: your agent can ask how it is put together and get answers that cite its files.

Show the connect command (Claude Code, Cursor, Codex, VS Code)

Run it in a terminal, then type /mcp in Claude Code and sign in with Repowise. No key to paste.

Docs
bash
claude mcp add --transport http repowise https://api.repowise.dev/mcp/deepseek-ai/dsh-superagent-prototype
Claude Code, Claude, Cursor and VS Code sign you in on first use, free. Codex and Windsurf need an API key: sign in to create one.Sign in for a key

This repo is public, so anyone with a free Repowise account can connect.

Try asking

  • get_overview()
  • get_answer("how is dsh-superagent-prototype put together?")
  • get_why("why is dsh-superagent-prototype structured this way?")

Ask about dsh-superagent-prototype

Answers come from dsh-superagent-prototype's README and source, with the files they cite. Two free answers a day without an account.

Want this for your own codebase? Map your repo free

About this page

Get a page like this for your own codebase

Repowise built this map of deepseek-ai/dsh-superagent-prototype from its source and git history. It parses every symbol, draws the dependency graph, scores each file's health from complexity, duplication, coverage and churn, finds hotspots and owners in the history, and writes documentation for every module. Your coding agent reads the same index over MCP, so its answers cite real files.

Free for 2 public repos, with a page like this one. No card. See how Repowise works

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed. Star Repowise on GitHub

deepseek-ai/dsh-superagent-prototype FAQ

What is dsh-superagent-prototype?

dsh-superagent-prototype is a research prototype of a proactive coding agent for one GitHub repository. It is packaged as a bundle for DeepSeek Harness and is not an official DeepSeek product.

How does dsh-superagent-prototype work?

dsh-superagent-prototype listens to GitHub and optionally Feishu for requests and opens a work item for each one. It runs the work in its own agent session and directory, then answers in the original conversation, for example with a pull request.

What language is dsh-superagent-prototype written in?

dsh-superagent-prototype is written mainly in TypeScript, with one Python file in the repository. Its dependencies are managed with npm.

How is the dsh-superagent-prototype code organized?

The src/app part turns incoming work into records and queues follow-up work, and src/core handles the lifecycle of each ticket. The src/client part contains the browser panel for watching tickets, sources, and settings.

Who is dsh-superagent-prototype for?

dsh-superagent-prototype is for teams that want an agent to act on one GitHub repository only when asked or when a team rule allows it. Its README calls it research code, so it should be reviewed and adapted before use in your own environment.

What license does dsh-superagent-prototype use?

dsh-superagent-prototype is released under the MIT license, identified as MIT.

Code health

8.7out of 10Excellent

On the leaderboards:#8 of 59 TypeScript repos

Key features of dsh-superagent-prototype

  • Listens to GitHub and, optionally, Feishu for requests from people who are allowed to ask for work.
  • Opens a work item for each request and works on it in its own agent session and directory.
  • Replies in the original conversation, for example with a pull request.
  • Provides a web panel that shows which items need a person.
  • 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.
  • Includes SETUP.md, which a coding agent can follow with the operator to install and configure the project.

How deepseek-ai/dsh-superagent-prototype works

Trigger events enter through the sources layer, which receives requests from GitHub and Feishu. The src/app part turns each trigger into a card, builds the prompt text the sessions run under, and queues the follow-up work that delivers results. The src/core part starts, runs, and ends work on each ticket, taking in triggers from sources, reports from workers, and decisions from assessors, and returning updated state with queued messages and domain events. The src directory wires the executor that carries out ticket work into the host application, and src/index.ts, src/app/tools/main.ts, and src/client/app.tsx are the entry points. src/client/app.tsx is the browser side of the agent panel, where operators watch tickets, sources, and settings. docs/000-philosophy.md explains the principle that the agent acts only when asked or when a team rule allows it.

The codebase has 206 files and 1,784 symbols in 4 modules, led by TypeScript and Python. Code health is 8.7 out of 10, rated excellent.