#164164th of 197 in Coding agents and AI editors
dsh-superagent-prototype
Agent that takes GitHub requests for one repo and replies with pull requests
- Stars
- 22.4k
- License
- MIT
- Last commit
- Oct 2026
- Language
- TypeScript
494 on 26 Oct 202622,427 on 9 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
dsh-superagent is a coding agent for a single GitHub repository, packaged as a bundle for DeepSeek Harness. It listens to GitHub and optionally Feishu, opens a work item when an allowed user asks, works in its own agent session and directory, and answers in the original conversation, for example with a pull request. A web panel lists items that need a person.
Who it is for: Teams exploring proactive, repo-level coding agents on DeepSeek Harness
Strengths
- Acts only on request or team-written rules; authorization is documented in docs/002
- Each work item runs in its own agent session and directory
- Numbered docs cover domain model, execution, prompts, panel and deployment
- Cookbook documents adding a new event Source
Weaknesses
- Labeled a research prototype, not production-ready, and may contain bugs
- Requires DeepSeek Harness as the runtime
- Scoped to one GitHub repository per deployment
- Some designs from the linked article are not implemented
What it needs
- Needs DeepSeek Harness, GitHub, Feishu (optional)
Quick answers
Is dsh-superagent-prototype open source?
Yes. Its code is on GitHub under the MIT license.
Also in Coding agents and AI editors
See all 197| Rank | Project | Score |
|---|---|---|
| 1 | claw-codeRust CLI coding agent harness using Anthropic and OpenAI-compatible APIs | 97 out of 100 |
| 2 | t3codeDesktop, web and mobile front end for controlling local coding agent CLIs | 94 out of 100 |
| 3 | codewhaleTerminal coding agent in Rust that works with 40+ model providers | 94 out of 100 |
| #4 | piMinimal terminal coding agent extensible with skills and packages | 92 out of 100 |
| #5 | oh-my-openagentTerminal coding agent with multi-model orchestration and git-backed memory | 92 out of 100 |