deepseek-harness vs conductor
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
conductor
Durable workflow engine that orchestrates microservices and LLM agents from JSON
Verdict
deepseek-harness ranks higher, 98 to 91. Of the 4 score parts we measure for both, deepseek-harness leads on 3, conductor on none and 1 is even.
- Adoptiondeepseek-harness leads by 7 points, 100 to 93.
- FreshnessEven at 100.
- Maintenancedeepseek-harness leads by 14 points, 100 to 86.
- Agent-readydeepseek-harness leads by 15 points, 85 to 70: 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 conductor if you are
Backend teams needing durable orchestration for services and AI agent loops
- Every step is persisted, with retry, restart, rerun, pause and terminate controls on executions
- Workflows are versioned JSON; running executions stay pinned to the version they started on
| What we compare | deepseek-harness | conductor |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 100, widely used | 93, widely used |
| Momentum | Not scored for this kind | Not scored for this kind |
| Freshness | 100, active | 100, active |
| Maintenance | 100, healthy | 86, healthy |
| Agent-ready | 85, ready | 70, partly |
| Facts from GitHub and the README | ||
| Stars | 278.8k | 32.5k |
| License | MIT (permissive) | Apache-2.0 (permissive) |
| Last commit | Nov 2026 | Nov 2026 |
| Last release | None published | Oct 2026 |
| Language | TypeScript | Java |
| Docker | No | Yes |
| GPU | Not stated | Not stated |
| arm64 or Apple Silicon | Not stated | Mentioned |
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
- Safety notice must be read before running; its contents are not in the README
- TypeScript
- Needs Node.js, pnpm (source build)
conductor
Conductor is a Java workflow orchestration server, originally from Netflix, that persists every task and runs versioned JSON workflow graphs. Workers in Java, Python, Go, JavaScript, C#, Ruby or Rust poll for tasks, while built-in tasks cover LLM chat, MCP tool discovery and calls, human approval, loops, forks and sub-workflows. It runs on Redis, Postgres or MySQL, with Elasticsearch or OpenSearch for indexing.
Who it is for: Backend teams needing durable orchestration for services and AI agent loops
Strengths
- Every step is persisted, with retry, restart, rerun, pause and terminate controls on executions
- Workflows are versioned JSON; running executions stay pinned to the version they started on
- Native LLM_CHAT_COMPLETE, LIST_MCP_TOOLS and CALL_MCP_TOOL tasks for agent loops
- Multiple persistence backends (Redis, Postgres, MySQL) and polyglot worker SDKs
Weaknesses
- Requires Java 25+ and a separate index store such as Elasticsearch or OpenSearch
- Orchestration is declarative JSON, not code, which suits some teams less than code-first engines
- Ruby and Rust SDKs are marked incubating
- Not a ready-made agent app; you build agents by composing workflow tasks
- Java, OpenAI (via llmProvider)
- Needs Java 25+, Node.js 16+ (CLI), Redis, Postgres or MySQL, Elasticsearch or OpenSearch
- Docker
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.