DeepSeek Harness

DeepSeek's open-source, plugin-based agent harness (dsh) with a local Web UI, a Python SDK and support for DeepSeek and third-party models.

Category

Harnesses

Website

https://github.com/deepseek-ai/deepseek-harness

Repository

https://github.com/deepseek-ai/deepseek-harness

Docs

https://deepseek-harness.github.io/deepseek-harness/

Status

beta

Pricing

open-source

License

MIT

Maintainer

DeepSeek

Added

2026-10-31

Updated

2026-10-31

Last verified

2026-10-31

Submitted by

agentica-curator

Tags

• coding-agents
• plugins
• web-ui
• python-sdk
• deepseek

Description

Built on an "everything is a plugin" architecture using the Cordis framework. Run npx @deepseek-ai/dsh web to start the Web UI at http://127.0.0.1:3080, or build from source with pnpm. Models are configured in the UI: DeepSeek via an API key, plus shipped third-party providers (e.g. anthropic, openai, moonshotai, zai) or any custom model API. Includes a Python SDK (pip install deepseek-harness-sdk), an MCP client package with reference memory-MCP configs, and plugin markets; plugins are tagged dsh-plugin on GitHub. MIT licensed. It is a developer preview with breaking changes expected; the project's SAFETY.md says it has not been security-audited and should run with least privilege, ideally in a disposable VM or container. The npm package is at 0.2.1-rc.5 (Oct 2026).

Agent access

llms.txt

https://deepseek-harness.github.io/deepseek-harness/llms.txt

Auth

api-key

Notes

Needs a model provider API key (DeepSeek or another provider) entered in Settings → Models.

Related

• Claude Code: Anthropic's agentic coding tool that reads your codebase, edits files, runs commands and integrates with dev tools; available in terminal, IDE, desktop and browser.
• OpenAI Codex CLI: OpenAI's lightweight open-source coding agent that runs in your terminal.
• OpenCode: The open-source AI coding agent, built for the terminal.

Guides & comparisons

• Research memory libraries compared: A-MEM vs MemoryOS vs MemOS vs LightRAG (Comparison): Four open-source memory systems that started as research papers. What each one stores, how many LLM calls a write costs, whether memories survive a restart, and how close each is to something you can run in an agent today.

Sources

• https://github.com/deepseek-ai/deepseek-harness
• https://deepseek-harness.github.io/deepseek-harness/
• https://www.npmjs.com/package/@deepseek-ai/dsh
• https://github.com/deepseek-ai/deepseek-harness/blob/master/SAFETY.md
• https://deepseek-harness.github.io/deepseek-harness/en/guide/providers.md
• https://deepseek-harness.github.io/deepseek-harness/en/guide/python-sdk.md

Machine-readable

• JSON
• Markdown
• Source file on GitHub (edit via pull request)