deepseek-harness vs adk-recipes
Two of the top projects in Agent backends, 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
adk-recipes
Runnable Agent Development Kit recipes, from single patterns to deployable agents
Verdict
deepseek-harness ranks higher, 70 to 52. Of the 5 score parts we measure for both, deepseek-harness leads on 3, adk-recipes on none and 2 are even.
- Adoptiondeepseek-harness leads by 32 points, 100 to 68: a wide gap.
- FreshnessEven at 100.
- Maintenancedeepseek-harness leads by 36 points, 100 to 64: a wide gap.
- Easy to runEven at 0.
- Agent-readydeepseek-harness leads by 55 points, 85 to 30: 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 adk-recipes if you are
Teams standardizing on ADK and Google Cloud
- Core recipes isolate one pattern each, so they lift cleanly into your project
- Vertical agents include deploy paths to Agent Engine and Cloud Run
| What we compare | deepseek-harness | adk-recipes |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 100, widely used | 68, popular |
| Freshness | 100, active | 100, active |
| Maintenance | 100, healthy | 64, fair |
| Easy to run | 0, hard | 0, hard |
| Agent-ready | 85, ready | 30, minimal |
| Facts from GitHub and the README | ||
| Stars | 249.7k | 10.5k |
| License | MIT (permissive) | Apache-2.0 (permissive) |
| Last commit | Oct 2026 | Oct 2026 |
| Last release | None published | None published |
| Language | TypeScript | Python |
| 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)
adk-recipes
A recipe collection for Google's Agent Development Kit: core/ holds single-pattern agents (OAuth flows, session memory, guardrails, RAG), contrib/ holds deployable vertical agents targeting Agent Engine, Cloud Run and Gemini Enterprise, and plugins/ holds skills packaged with SKILL.md and EVAL.yaml. Each recipe has its own README; ADK SDKs exist for Python, TypeScript, Go, Java and Kotlin. For teams standardizing on ADK and Google Cloud.
Who it is for: Teams standardizing on ADK and Google Cloud
Strengths
- Core recipes isolate one pattern each, so they lift cleanly into your project
- Vertical agents include deploy paths to Agent Engine and Cloud Run
- Recipe checklist and handbook define a contribution standard; tests and AGENTS.md included
- Committed within the last day
Weaknesses
- Root README is an index; no single app, no shared setup
- Gemini and Google Cloud are the assumed model and deploy target
- README states recipes are demonstrations, not for production use
- Mixed languages and maturity across folders
- Python, google, google-adk
- Needs google-adk, google-api-key-or-vertex-ai
More in Agent backends
Building with these?
We help teams pick an open-source AI stack and run it in production. Client work never changes a rank.