deepseek-harness vs mcp-use
Two of the top projects in Agent and MCP starters, side by side: who leads on each score part and who each one suits, then setup, license, activity and what each review found.
deepseek-harness
Plugin-based agent harness from DeepSeek with a local Web UI
mcp-use
TypeScript framework for building MCP servers with React Views and an Inspector
Verdict
deepseek-harness ranks higher, 70 to 59. Of the 5 score parts we measure for both, deepseek-harness leads on 3, mcp-use on none and 2 are even.
- Adoptiondeepseek-harness leads by 22 points, 100 to 78: a wide gap.
- FreshnessEven at 100.
- Maintenancedeepseek-harness leads by 3 points, 100 to 97: close to even.
- Easy to runEven at 0.
- Agent-readydeepseek-harness leads by 85 points, 85 to 0: a wide gap.
Pick deepseek-harness if you are
Developers building agents through plugins, tolerant of breaking changes
- Starts with one npx command; no Docker or database mentioned
- Everything-is-a-plugin design on Cordis, with a dsh-plugin topic for discovery
Pick mcp-use if you are
TypeScript developers building MCP servers and ChatGPT or Claude apps
- Zod schemas flow from tool input and output into React View props
- Inspector served at /mcp/inspector in dev; headless CLI can call tools and screenshot Views
| What we compare | deepseek-harness | mcp-use |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 100, widely used | 78, popular |
| Freshness | 100, active | 100, active |
| Maintenance | 100, healthy | 97, healthy |
| Easy to run | 0, hard | 0, hard |
| Agent-ready | 85, ready | 0, none |
| Facts from GitHub and the README | ||
| Stars | 254.5k | 10.8k |
| License | MIT (permissive) | MIT (permissive) |
| Last commit | Oct 2026 | Oct 2026 |
| Last release | None published | Oct 2026 |
| Language | TypeScript | TypeScript |
| Docker | No | No |
| GPU | Not stated | Not needed |
| arm64 or Apple Silicon | Not stated | Not stated |
deepseek-harness
DeepSeek Harness (`dsh`) is a TypeScript agent harness built on the Cordis plugin framework, where every component is a plugin. It runs via `npx @deepseek-ai/dsh web` or from a source checkout with pnpm, and serves a Web UI at 127.0.0.1:3080 by default. It is a developer preview, and the README warns of compatibility-breaking changes.
Who it is for: Developers building agents through plugins, tolerant of breaking changes
Strengths
- Starts with one npx command; no Docker or database mentioned
- Everything-is-a-plugin design on Cordis, with a dsh-plugin topic for discovery
- MIT license, documentation site and third-party license notices
- Web UI binds to 127.0.0.1 by default; --no-open skips the browser
Weaknesses
- Developer preview; README warns of compatibility-breaking changes
- No tagged release published
- README does not list supported models, providers or system requirements
- Safety notice must be read before running; its contents are not in the README
- TypeScript
- Needs Node.js, pnpm (source build)
mcp-use
mcp-use is a TypeScript v2 framework, built on the official MCP TypeScript SDK, for writing MCP servers with Zod-typed tools. It binds React Views to tools so they render as interactive apps in hosts such as ChatGPT and Claude. It ships a dev Inspector, a headless client CLI, a View screenshot command, tunneling, and a deploy command targeting Manufact.
Who it is for: TypeScript developers building MCP servers and ChatGPT or Claude apps
Strengths
- Zod schemas flow from tool input and output into React View props
- Inspector served at /mcp/inspector in dev; headless CLI can call tools and screenshot Views
- Built-in tunnel (mcp-use dev --tunnel) for testing in ChatGPT and Claude before deploy
- MIT licensed, with a self-hosting guide alongside the hosted option
Weaknesses
- v2 is a breaking change from v1; a migration guide is needed
- Benchmark and comparison figures are the project's own, not independent
- One-command npm run deploy targets the vendor's Manufact platform
- Example apps live in separate repos, not in the main package
- TypeScript
- Needs Node.js, npm, React, Zod
More in Agent and MCP starters
Building with these?
We help teams pick an open-source AI stack and run it in production. Client work never changes a rank.