A DeepSeek Harness plugin for persistent multi-model agent teams: per-member model/tool policy, bounded DAG execution, Run Center and token insights.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-08-23
License
MIT
01What can it help you accomplish?
Run persistent, reusable multi-model agent teams inside DeepSeek Harness
A saved team selected beside the normal composer; the lead model plans work, runs a bounded dependency graph, and synthesizes the result, with Run Center and Token insights
DeepSeek Harness (dsh) users who want to orchestrate several models/roles as one reusable team rather than one-off dispatches
02How to install into DeepSeek Harness
Not specified by the author
03DSH integration and capability boundaries
A dsh plugin that adds persistent, reusable multi-model agent teams to DeepSeek Harness with a Settings → Teams UI, Run Center and Token insights
Per-member model & tool policy
A team definition (members, roles, fallback routes, token limits, tool policies)→Combine a planner, implementer, reviewer or specialist without forcing one route on everyone
Runs a bounded DAG with retries, a quality gate and background work (observable, cancellable, restart-safe)Run Center & Token insights
Running and past team runs→Official-provider token usage (input, cache-read, …) and a run center to manage teams
04Who is it for? When not to use it?
Good for
- DeepSeek Harness (dsh) users who want to orchestrate several models/roles as one reusable team rather than one-off dispatches
05Compatibility, maintenance and safety notes
- The plugin offers Team / Solo / Inherited modes — a durable conversation override, a project default, or a one-message exception — to avoid ambiguous switches.
MIT · actively maintained (latest release v1.0.1, 2026-08-23)
06Frequently asked questions
What does dsh-agent-team-gui add to DeepSeek Harness?
Persistent, reusable multi-model teams. You create a team once in Settings → Teams, give each member its own model/role/fallback/token limit/tool policy, and select it beside the composer; the lead plans and runs a bounded DAG, then synthesizes.
Is it a native DeepSeek Harness plugin?
Yes — it is built for DeepSeek Harness and adds a Teams UI, Run Center and Token insights. The README excerpt we captured shows the feature overview but not a verbatim dsh install command, so we do not assert specific steps.
How are long runs kept safe?
The plugin uses a bounded DAG with retries, a quality gate and background work — observable, cancellable, finite and restart-safe.
Can I mix a team with a normal session?
Yes. It offers Team / Solo / Inherited modes: a durable conversation override, a project default, or a one-message exception, so you avoid ambiguous switches.
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.
petdex
by crafter-station
A public gallery of animated pets for Codex, Claude Code, DeepSeek Harness, Hermes, OpenCode, Gemini CLI, and more.
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
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
**Persistent, reusable multi-model Agent teams for [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness).*…
Give each member its own model, role, fallback route, token limit, and tool policy.
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-08-23. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.