Skip to content
NewA calmer, faster ZeroHour

DeepSeek Open-Sources dsh-superagent Router-Worker Coding Agent

Original title: DeepSeek Open-Sources dsh-superagent, a Router-Worker Coding Agent

MarkTechPostby Asif Razzaq ()7 min read1
AI summary · x-ai/grok-4.7Summaries by Vani AI

DeepSeek releases dsh-superagent a proactive router-worker coding agent for GitHub repositories.

DeepSeek open-sourced dsh-superagent an MIT-licensed research prototype. The agent uses a router-worker architecture to handle multiple coding tasks in parallel sessions. It listens to GitHub and Feishu events turning them into isolated worker sessions for faster task completion.

Key facts
  • p95 task completion 15.7 minutes vs 29.5 minutes baseline
  • 1.9x faster under noisy message streams
  • Runs as DeepSeek Harness profile on local machine
  • Uses Node.js 22+ git Python tmux and GitHub CLI
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

From the original article

7 min read at marktechpost.com

DeepSeek released dsh-superagent, an MIT-licensed research prototype that turns GitHub and Feishu requests into parallel worker sessions. In DeepSeek’s own simulation, p95 task completion fell to 15. 7 minutes from 29.

MarkTechPost

By Asif Razzaq · marktechpost.com

Read the full article at MarkTechPost

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.