DeepSeek Harness
- ID
- 13902
- Status
- summarized
- Published
- 13 Aug 2026, 8:58 PM
- Fetched
- 14 Aug 2026, 12:41 AM
- Provider
- Hacker News
- Category
- dev-community
- Original URL
- https://github.com/deepseek-ai/deepseek-harness
- Source URL
- https://hnrss.org/best
Summary
- Score
- 7.0
- Created
- 14 Aug 2026, 12:41 AM
- Tags
- Audience
- developersvibe_codersai_ml_learnersai_agent_users
What happened
DeepSeek AI released DeepSeek Harness (dsh), an open-source agent harness with a plugin-first architecture powered by Cordis, described in a paper on spatiotemporal composability. It is in developer preview with explicit warnings of compatibility-breaking changes, and can be launched via `npx @deepseek-ai/dsh web` (Web UI at port 3080). The repo has 33.6k stars and 2.6k forks.
Why it matters
If you are building or evaluating AI agent tooling, dsh offers a plugin-extensible harness you can try locally with one npx command, but the explicit breaking-change warning means you should treat it as experimental and avoid production dependencies until it stabilizes.
Discussion angle
Compare dsh's 'everything is a plugin' architecture against other agent harnesses the audience may already use—what does plugin-first composability buy you, and is the breaking-change risk acceptable for your current projects?