mrkeyoor.comAsk

deepseek-ai/deepseek-harnesstested 13 Oct 2026commit d743267

deepseek-harness review

Our DeepSeek Harness build passed in 350 seconds, but its tests timed out at 900 seconds, so we recommend a contained developer evaluation rather than a production commitment.

  • Build 350 s
  • Tests timed out
36 sto install
  • 1,397 packages
LLM Toolsevaluationupdated 14 Oct 2026
+2,748stars / 7d

DeepSeek Harness is an agent application that lets an AI model read and edit project files, run commands, and work through tasks in a browser or other interfaces. Its main appeal is that developers can replace the agent's internal parts through plugins instead of maintaining a separate fork.

Verdict

Our DeepSeek Harness build passed in 350 seconds, but its tests timed out at 900 seconds, so we recommend a contained developer evaluation rather than a production commitment. Use it when changing the agent's internal behavior is a requirement and you can absorb breaking plugin updates. Choose a more direct coding assistant when repository editing is all you need.

We ran it

Lab card: what happened when we ran deepseek-harness
Lab card from our sandbox run
Screenshot of deepseek-harness (deepseek.com/harness)
deepseek.com/harness
Install✓ · 36s1397 packages · 2065 MB
Build✓ · 350s
Tests✗ timed out · 900sran, no count parsed
Repo14899 files~1,214,194 lines of source · 126.3 MB · 23 CI workflows

Answers from our run

Does deepseek-harness build from source?

Dependencies installed in 36 seconds (1397 packages), and the build succeeded in 350 seconds. We cloned commit d743267 into a clean Debian container with 3 CPUs and no project-specific setup.

Do deepseek-harness's tests pass?

We could not finish them: the suite was still running after 15 minutes in our container.

Who should not use deepseek-harness?

Teams requiring stable plugin contracts: the README explicitly warns of compatibility-breaking changes during developer preview.

What are the alternatives to deepseek-harness?

Aider, OpenCode. Our DeepSeek Harness build passed in 350 seconds, but its tests timed out at 900 seconds, so we recommend a contained developer evaluation rather than a production commitment.

Setup2/52,065 MB installed, 350-second build, and an incomplete suite
Docs4/5Detailed plugin, provider, platform, and safety guidance
Community3/5Recent push and Discussions support, no issue activity to assess
Maturity2/5Developer preview warns of breaking changes and lacks a security audit

Discussed on

  1. hnDeepSeek harness: what doors does config over code open?3 points

Who it’s for

Agent developers who need to replace model adapters, tools, or the agent loop through Cordis plugins.
Teams testing a custom model gateway and willing to adjust request compatibility in YAML.
Contributors who can work in a large pnpm monorepo and accept developer-preview changes.

Who it’s NOT for

Teams requiring stable plugin contracts: the README explicitly warns of compatibility-breaking changes during developer preview.
Operators depending on the agent as their security boundary: SAFETY.md says it has not had a security audit and must not be treated as secure or production-ready.
Users expecting OAuth sign-in through the Models settings: the provider guide says OAuth providers are unsupported there.
Contributors requiring a completed full suite on a modest machine: our tests reached the 900-second limit without a final result.

Setup reality

Our sandbox installed commit d743267 in 36 seconds, pulling 1,397 pnpm packages and using 2,065 MB on disk. The build passed in 350 seconds. Tests timed out at 900 seconds, with passing package-level checks still visible in the tail and no final suite summary.

The Web UI needs a configured model and selected workspace before you can submit a task. The documented DeepSeek route needs an API key. Custom gateways need their endpoint, protocol, credentials, and model entries, with some compatibility settings edited in cordis.patch.yml. OAuth sign-in is unsupported in the Models settings.

Source development needs Node's TypeScript stripping support and Corepack-enabled pnpm. Keep Windows or WSL dependencies within their own operating system environment. The README labels the project developer preview, and the safety notice recommends disposable environments because permissions do not guarantee isolation.

Replace the agent loop when the default gets in your way

DeepSeek Harness lets you replace the part of an agent that decides what happens next. Its Cordis architecture makes the model adapter, tool registry, session log, and agent loop plugins. That matters if your team needs a different approval policy or execution environment without maintaining a fork of the whole application. An ordinary developer who just wants help editing a repository has less reason to learn this machinery.

The source checkout contained 14,899 files and roughly 1,214,194 lines in our lab's source scan. Those figures put some weight behind the word "harness": adopting the development checkout means entering a substantial TypeScript monorepo. The architecture reference explains how plugins register services and events, then undo their registrations when unloaded. Read that before deciding to write your own extension. The attraction is control over behavior, with a sizeable codebase attached.

What happened when we ran it

Our run installed 1,397 pnpm packages in 36 seconds, used 2,065 MB on disk, and built successfully in 350 seconds. We cloned commit d743267 on October 13, 2026 into a fresh Debian container with 3 CPUs and 8 GB of RAM, using lab-node:22. The container was unprivileged and held no secrets. Installation and compilation both completed, which supports trying the source route on a comparable machine.

The tests timed out at 900 seconds. The supplied tail still shows passing checks for session search, filesystem integration, pending submissions, and YAML reload behavior. Other passing checks concern preserving a user's pre-existing uncommitted work and recording changes after tools settle. The tail contains no final suite summary or exception identifying the reason for the timeout. We cannot call the whole suite green, invent a total passed count, or blame a missing system package.

The scan found 23 CI workflow files and monorepo workspaces, with no Dockerfile or top-level tests directory. That last signal needs care: the log explicitly names test files inside package directories. Tests exist, and some ran successfully. The useful finding is the incomplete full run. For a team contributing changes, budget time to identify the relevant checks and establish a repeatable local result before relying on the whole suite as a release condition.

A running browser still needs a model and workspace

The 350-second source build prepares artifacts that pnpm dsh web reuses. The README also offers npx @deepseek-ai/dsh web, which launches a browser interface on loopback. We measured the checkout path, so the npm shortcut carries no timing claim here. Once the browser opens, the Web UI guide tells you to save a DeepSeek API key and choose a workspace. The composer stays unavailable until that directory is selected.

Other providers and custom model endpoints are supported through the Models settings. A custom endpoint needs a provider identifier, base URL, protocol, credential, and model entry. The provider guide warns that a correct key and reachable gateway can still produce rejected requests when their request formats differ. Some compatibility settings require editing cordis.patch.yml. Keys live in the Harness home's credentials file, and OAuth providers are unsupported in this settings flow.

Source development carries a large rebuild cost

A 126.3 MB checkout became a 2,065 MB installed environment in our sandbox. Keep that footprint in mind before choosing the source tree for a disposable machine or shared development image. The development guide also requires Node's TypeScript stripping support and a Corepack-enabled pnpm setup. Disabling that runtime capability can prevent the build scripts from loading their configuration. The README's short command sequence leaves these details to the contributor documentation.

Windows and WSL development have their own instructions: keep the checkout, dependencies, and toolchain in the same operating system environment. The guide warns about cross-filesystem overhead and platform-specific binaries. Our 350-second Debian build establishes nothing about a Windows workstation's experience. If Windows is your deployment target, validate that path separately rather than treating a Linux build as portable proof. The documentation earns credit for spelling out this boundary.

Developer preview excludes a production security promise

The 900-second timeout is one reason to qualify adoption, and the project's own warning is another. Its safety notice says the software has not undergone a security audit and must not be treated as secure or production-ready. Model-generated commands and third-party plugins can reach whatever files, credentials, processes, and network access you provide. Permissions and approval prompts reduce exposure without guaranteeing isolation. Choose a disposable environment and narrowly scoped access for an evaluation.

The README also promises compatibility-breaking changes during developer preview. That rules out a comfortable recommendation for teams demanding stable extension contracts. GitHub recorded the last push on October 9, 2026 and listed zero open issues and PRs, while the latest-release endpoint returned no release. Feedback goes through Discussions, so the empty issue list provides no evidence that users have no problems. The recent push supports ongoing work, though issue activity cannot establish support responsiveness here.

Pick a coding assistant unless you need to change its internals

DeepSeek Harness's 1,397-package source setup makes most sense when replacing adapters or execution behavior is part of your job. Aider is a better starting point for terminal pair programming with Git-managed edits. OpenCode suits someone who wants a coding agent with separate build and read-only planning modes. Neither comparison carries a speed claim. Try Harness when you can name the plugin boundary you need to change and accept maintaining that choice through breaking updates.

Alternatives

ProjectWhat it isPick it when
Aider gh↗A terminal pair-programming assistant with Git-managed edits.pick this instead when editing an existing repository is the job and replacing an agent loop is not.
OpenCode gh↗An open coding agent with build and read-only planning modes.pick this instead when you want a ready coding workflow rather than a Cordis plugin architecture.

What people are saying

  1. [velocity-scout] Zhou-Yujing114514/deepseek-harness-linux
  2. [velocity-scout] vobl1999/Deepseek-Harness-Easy-Setup

Sources

  1. DeepSeek Harness README
  2. Harness architecture
  3. Web UI setup
  4. Model configuration and limitations
  5. Safety notice
  6. Source development prerequisites
  7. GitHub repository facts
  8. GitHub release availability

More llm tools reviews