DeepSeek Harness plugin

dsh-kernel-mesh

Host-plane kernel mesh for DeepSeek Harness: kimi/grok/codex/minimax kernel adapters, L2 subagent recipes, and kernel tools.

Jump to install

Source facts

Repository
oppnc/dsh-kernel-mesh
Latest update
Aug 20, 2026
Category
Workflow & Automation
GitHub stars
0
Format
plugin
Catalog evidence
Upstream dsh.bundle evidence
Evidence path
package.json#dsh.bundle
Checked against
0.1.0-rc.8
Upstream check date
2026-08-20

This evidence comes from the upstream catalog. This site has not installed, run, or security-reviewed the plugin.

Install

Start with a prompt that asks an agent to review the GitHub repository and source. Switch to the command if you want to install it yourself.

Copy this prompt into DSH, Codex, or another agent and ask it to review the GitHub repository and source first.

Do not install or run any commands yet. Read this plugin's GitHub repository, README, and relevant source code. Then answer the questions below clearly and directly so I can decide whether it fits my needs:

1. What is this plugin, and what problem does it solve?
2. Who is it for, and what are its typical use cases?
3. How is it used after installation? Include one minimal example.
4. What known limitations or privacy, security, compatibility, or maintenance risks does it have?
5. Give a clear recommendation: recommend, conditionally recommend, or do not recommend, with reasons.

Distinguish statements documented by the repository, inferences from source code, and unknowns. If evidence is insufficient, say so explicitly. Do not guess or simply repeat the README.

GitHub: https://github.com/oppnc/dsh-kernel-mesh
Plugin: dsh-kernel-mesh
Author: oppnc

Check the source files

Read the README and other files from this plugin directory before installing.

File explorer4 files
README.mdSource · read only
README language

English | 中文

dsh-kernel-mesh

DSH runs on one simple idea: everything is a plugin. Models, tools, subagents — plug them together however you like.

So we did exactly that: we turned four coding harnesses — Kimi Code, Grok Build, OpenAI Codex, and MiniMax Mini-Agent — into DSH plugins, bundled in this one package.

The payoff is simple: switch to a Kimi / Grok / Codex / MiniMax model inside DSH, and it's no different from opening each one's own CLI. Every model stays in the environment it knows best — main agent or subagent, it feels like coming home.

What's inside

  • L1 kernel routeskimi-kernel / grok-kernel / codex-kernel / minimax-kernel, registered as DSH model routes your main agent can switch to.
  • L2 subagent recipeskimi-agent / kimi-explore / kimi-plan, grok-agent / grok-explore / grok-plan, codex-agent / codex-explore / codex-worker (each kernel's own subagent types; minimax has none upstream). Each recipe carries the vendor's own subagent prompt and tool whitelist, so a kernel subagent sees and uses exactly what that harness's subagent would — independent of the parent's preset.
  • Three kernel toolskernel_status, kernel_run, kernel_switch.
  • Vendor search tools, offered only when opted inkimi_search / kimi_fetch appear only if dsh-kernel-kimi is installed and a Moonshot credential exists; grok_search / grok_fetch appear only if dsh-kernel-grok is installed and a Grok OAuth credential exists. They stay separate tools (different corpora). Official web_search is DeepSeek's own search and is not a wrapper.

Kernel matrix

KernelWireEndpoint
kimi-kernelAnthropic Messageshttps://api.kimi.com/coding/v1/messages (kimi-cli 1.49.0: max_tokens clamped to remaining context; catalog includes k3)
grok-kernelResponses (proxy)https://cli-chat-proxy.grok.com/v1/responses
codex-kernelResponses (custom)your codex base_url + /responses
minimax-kernelAnthropic Messages (CN)https://api.minimaxi.com/anthropic/v1/messages

System prompts

Each kernel plugin registers the vendor's upstream system prompt (tool names and runtime placeholders adapted to the DSH tool surface) as the agent's sole system-prompt section (complete: true + suppressRuntimeContext()), so a session on a kernel sees ONLY that harness's prompt — not DSH's.

Fallback routes (opt-in)

The vendors have no fallback of their own, so by default each kernel uses its official API. If a kernel's own API is out of quota, set DSH_KERNEL_USE_FALLBACK=1 to route the kernels through your existing subscriptions instead:

KernelFallback routeModel
kimi-kernelollamakimi-k2.7-code
grok-kernelollamagpt-oss:120b
codex-kernelopencode-gogpt-5.6-luna
minimax-kernelollamaminimax-m3

Keys are read from ~/.dsh/.credentials.yaml (OLLAMA_API_KEY, MY_OPENCODE_GO_API_KEY). Leave DSH_KERNEL_USE_FALLBACK unset (or 0) for the official kernel APIs.

Install

Install the bundle into your profile with the official plugin command:

dsh plugin --profile web add github:oppnc/dsh-kernel-mesh

dsh plugin forwards to pnpm and reconciles dsh.profile.bundles automatically — this package declares "dsh": { "bundle": { "patch": "./cordis.patch.yml" } }, so it joins the profile's layer stack. Restart the profile afterwards:

dsh web

Usage

kernel_status                          # which kernels, L2 types, transports are registered
kernel_run(kernel, type, task)         # fan a task out to a foreign kernel
kernel_switch('kimi')                  # set the default model route for future sessions
Kerneltype values
kimicoder, explore, plan
grokgeneral, explore, plan
codexexplore, worker
minimax(none — no subagent tool upstream)

Screenshots

The whole family on GitHub — every README has a one-tap language switch:

<table> <tr> <td><img src="docs/screenshots/01-mesh-readme-en.png" alt="dsh-kernel-mesh README (EN)" width="410"></td> <td><img src="docs/screenshots/02-mesh-readme-zh.png" alt="dsh-kernel-mesh README (中文)" width="410"></td> <td><img src="docs/screenshots/03-kimi-readme.png" alt="dsh-kernel-kimi README" width="410"></td> </tr> <tr> <td><img src="docs/screenshots/04-grok-readme.png" alt="dsh-kernel-grok README" width="410"></td> <td><img src="docs/screenshots/05-codex-readme.png" alt="dsh-kernel-codex README" width="410"></td> <td><img src="docs/screenshots/06-minimax-readme.png" alt="dsh-kernel-minimax README" width="410"></td> </tr> </table>

Full-size images live in [docs/screenshots/](docs/screenshots/).

License

MIT — see [LICENSE](LICENSE).