A native DSH web-ui plugin that adds custom model endpoints and per-model capability tuning to the Harness Web UI, with models.dev auto-fill and credential-store API Key handling.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-09-05
License
MIT
01What can it help you accomplish?
Create custom model endpoints for the DSH Web UI
Configurable endpoints (name/URL/API Key/protocol) supporting openai-completions, openai-responses and anthropic-messages, with per-model capability params
DSH Web UI operators who need custom or self-hosted model providers
Tune per-model capabilities with models.dev auto-fill
Saved llm-pi-ai model configs — context window, max output, input modalities and reasoningEfforts — with a secret-free config preview
Users who need precise per-model parameter control
02How to install into DeepSeek Harness
Prerequisites
- A running `dsh` (DeepSeek Harness) CLI installed
Installation steps
- 01
$ dsh plugin --profile web add github:MarvekG/deepseek-harness-model-config
- 02
$ dsh web
- 03
Open Web UI Settings → Model Advanced Configuration to add endpoints or edit llm-pi-ai configs
Verify the integration
Not specified by the author
03DSH integration and capability boundaries
Installed as a native DeepSeek Harness web-ui plugin via the official dsh CLI (`dsh plugin --profile web add github:...`); adds a Model Advanced Configuration page to the DSH Web UI.
Custom endpoints with provider headers
name, URL, API Key, protocol→endpoint supporting openai-completions / openai-responses / anthropic-messages
Custom request Headers applied at provider layer to model requestsUnified model discovery (GET /models)
endpoint URL + API Key→candidate models with search / select-all / invert / clear
models.dev auto-fill
model IDs / display names→filled context window, max output, modalities and reasoning from models.dev
Credential-store API Key handling
API Key→key written only to Harness credential store
API Key never enters settings.yaml or the config preview
04Who is it for? When not to use it?
Good for
- DSH Web UI operators who need custom or self-hosted model providers
- Users who need precise per-model parameter control
05Compatibility, maintenance and safety notes
- Model parameters auto-fill only when https://models.dev is reachable.
- Custom request Headers are used only for model requests; "Fetch available models" still uses the API Key.
- API Key is written only through the Harness credential store and never lands in settings.yaml or the config preview.
MIT · actively maintained (latest release 2.3.0, 2026-09-04)
06Frequently asked questions
How do I install deepseek-harness-model-config?
Run `dsh plugin --profile web add github:MarvekG/deepseek-harness-model-config` then `dsh web`. Open Settings → Model Advanced Configuration in the Web UI. Pin a version by appending `#<sha>` to the repo address.
Which model protocols does it support?
It supports `openai-completions`, `openai-responses` and `anthropic-messages` endpoints, configurable with a name, URL, API Key and protocol, plus custom request Headers at the provider layer.
Where is my API Key stored?
The API Key is written only through the Harness credential store; it never enters `settings.yaml` or the configuration preview shown before saving.
Why are some model parameters not auto-filled?
Auto-fill of context window, max output, input modalities and reasoning relies on https://models.dev being reachable. Without access, you can still type model IDs manually.
07Related DSH workflows
dsh-web
by zhu1090093659
One-command install of a full DSH web workbench: monitoring HUD, task board, SSH panel, mobile remote, vision tool, git graph, skins, plugin manager.
deepseek-balance-whale-widget
by meteornox
Floating cartoon whale widget living in the DSH Web UI corner that auto-loads, refreshes DeepSeek balance every 60s (click to refresh), with drag/snap, left-flip mirror, squish animation, size scaling and rolling-number animation
petdex
by crafter-station
A public gallery of animated pets for Codex, Claude Code, DeepSeek Harness, Hermes, OpenCode, Gemini CLI, and more.
j-space-cognition-suite
by tiger3807861189
J-Space Cognition Suite — a model-agnostic inference-time control suite for deep reasoning, long-horizon work, verification, and recovery. Based on Anthropic's J-space global workspace research.
08Data and sources
# DeepSeek Harness 模型配置插件
dsh plugin --profile web add github:MarvekG/deepseek-harness-model-config
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-09-05. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.