DeepSeek Harness Plugin Hub

Publish and manage complete Harness Profiles. Discover Plugins for your next setup.

Explore

PluginsPresetsDocsNews

Community

Publish a pluginContactReport an issue

Resources

Plugin Hub on GitHubDeepSeek HarnessSystem statusPrivacy notice
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

Independent and unofficial. Not affiliated with, authorized by, or endorsed by DeepSeek.

CATEGORY

Agents & Orchestration DSH plugins

Verified manifests and exact versions in this category.

All categories

All pluginsAgents & Orchestration1.5kMemory & Context2.1kCoding & Developer Tools4kUI & Customization5.6kIntegrations & Communication3kVision & Media916Search & Research1.2kSecurity & Access1.8kModels & Usage2.5kProductivity & Workflow3.6k

All plugins

1552 plugins
Pv0.1.1
dsh-plan-runner

Plan Runner

Step-by-step plan runner for DeepSeek Harness, built for small local models: each step runs in a fresh child conversation with its own model and access, verified by mechanical checks and an optional read-only reviewer; one retry, then a finer split proposed to the user.

↓ 0★ 0
Updated 10/25/2026
Lv0.2.0
dsh-live-docs

Live Docs

Automatically delivers changes to live documents at specified absolute paths to the model via tail appends. The mechanism mirrors native dsh-agent-instructions (four states / state included in source / framework escaping / budget degradation), changing only the anchors from the cwd ancestor chain to

↓ 0★ 0
Updated 10/25/2026
Ev1.0.0
@zilorn/dsh-efficient-mode

Efficient Mode

Efficient Mode (高效模式) agent preset with a base prompt interpolated from live git and environment facts

↓ 0★ 0
Updated 10/25/2026
v0.5.1
@hanphone/dsh-a2a

A2a

Agent2Agent (A2A) v1.0.1 dual-end plugin for DeepSeek Harness: multiple inbound servers (per-instance agent presets; skills derived from each preset's skill directory), multiple outbound connections, GUI-managed instance lifecycle.

↓ 0★ 0
Updated 10/25/2026
Sv0.2.1
dsh-session-mesh

Session Mesh

Cross-session mesh: list every session in every workspace, message any session as a real agent message, create/rename/archive/cancel sessions, and wait for session changes.

↓ 0★ 0
Updated 10/25/2026
Uv1.0.0
dsh-ultra

Ultra

Ultra mode for DeepSeek Harness: standing multi-agent orchestration and max reasoning effort. A DSH port of Claude Code's ultracode.

↓ 0★ 0
Updated 10/25/2026
Lv1.1.1
dsh-lan-relay

Lan Relay

DeepSeek Harness LAN relay plugin: one machine runs the hub within the DSH process, while other machines can join by pasting the join code; enables cross-machine Agent messaging, offline message redelivery, and automatic session wake-up. Standard DSH bundle, ready to use after installation.

↓ 0★ 0
Updated 10/25/2026
Sv0.1.0
dsh-structured-output

Structured Output

DSH plugin: Enforces a fixed output structure for Claude Code / DSH subagents (supports fixed-enumeration choice, open-choice options with confidence via open_choice, and fixed JSON). Implemented by enforcing tool calls through a proxy layer. Uses a single model, with a different schema configured f

↓ 0★ 0
Updated 10/25/2026
Sv0.1.0
dsh-sts2-agent

Sts2 Agent

Let an AI play Slay the Spire 2: screenshot + mouse/keyboard tools, with an automatic play log.

↓ 0★ 0
Updated 10/25/2026
v0.1.0
dsh-baseline

Baseline

A fail-closed, transparent baseline gate for DeepSeek Harness: network egress policy, monotonic subagent permission ceiling, and one verifiable decision/result audit chain.

↓ 0★ 0
Updated 10/24/2026
v0.3.0
@agi321/dsh-temp-chat

Temp Chat

Temporary Q&A: /single-turn and /multi-turn answer via an isolated web-researching subagent; results stay in the conversation and never enter the main model context.

↓ 0★ 0
Updated 10/24/2026
Rv0.2.0
dsh-readme

Readme

Read a user-designated text file before every agent's first step and inject its contents into the system prompt as binding working constraints.

↓ 0★ 0
Updated 10/24/2026
Cv0.1.0
dsh-creativity-nudge

Creativity Nudge

Advisory creativity trigger: inject a single divergent reminder when the model locks onto a path. Advisory creativity nudge for DSH (off by default).

↓ 0★ 0
Updated 10/24/2026
Sv2.1.0
dsh-staff

Staff

DSH Staff — Staff system: persistent staff roster, task ledger, group and private chat, turning Agent Teams into manageable, conversational, and reviewable staff teams.

↓ 0★ 0
Updated 10/24/2026
Qv0.1.0
qq-qzone-diary

Qq Qzone Diary

Let AI characters casually post QQ updates based on their own moods: random timing, random length, and images they choose themselves.

↓ 0★ 0
Updated 10/24/2026
Kv0.3.6
dsh-kakaka-chat

Kakaka Chat

"Conversation Mode" DeepSeek Harness Agent preset: conversation/creation mode based on the kakake AI assistant

↓ 0★ 0
Updated 10/24/2026
Av0.1.0
dsh-ai-relay

Ai Relay

Ask other AI models to help with the work: OpenAI-compatible and Anthropic routes as agent tools inside DeepSeek Harness.

↓ 0★ 0
Updated 10/24/2026
Av2.2.0
dsh-agent-personality

Agent Personality

A personality with a signed emotion variable (-10..+10), its own worldview, diary, relationship network and a voice that opens conversations. For DeepSeek Harness.

↓ 0
Updated 10/24/2026
Tv0.2.0
dsh-task-plan

Task Plan

DeepSeek Harness plugin: an off-peak task plan table that runs pending tasks as concurrent Sessions

↓ 0★ 0
Updated 10/24/2026
Cv1.4.0
@local/dsh-cradle

Cradle

Sovereign Agent Diplomacy with a pixel room and agency: every agent owns a 2D pixel room and an avatar it walks, acts and extends through tools, wakes on its own frame clock under a bounded daily budget, chats with the user, and sleeps when that budget is spent.

↓ 0★ 0
Updated 10/24/2026
Tv0.2.0
dsh-tide-whisper

Tide Whisper

潮声: lets the AI in DSH speak up on its own at irregular intervals—random intervals, random silences, and context and constraints injected when it speaks instead of fixed lines, with a toggle in the settings page

↓ 0
Updated 10/24/2026
Wv0.1.3
dsh-waibao

Waibao

Codex and DSH desktop parallel subtasks, communication during execution, free models, and continuous context.

↓ 0★ 0
Updated 10/24/2026
Cv0.1.1
dsh-channel-telegram

Channel Telegram

DSH bundle: thin Telegram bridge (long-poll Bot API, allowlist, per-chat agent sessions).

↓ 0★ 0
Updated 10/24/2026
Sv0.1.5
dsh-subagent-output-sanitizer

Subagent Output Sanitizer

DSH plugin dsh-subagent-output-sanitizer

↓ 0★ 0
Updated 10/24/2026
v1.0.1
@exagone313/dsh-subagent-sidebar

Subagent Sidebar

Subagent catalog tab for the DeepSeek Harness (dsh) right sidebar.

↓ 0★ 0
Updated 10/24/2026
Av0.1.0
ark-auto-check

Ark Auto Check

Skill plugin for DeepSeek Harness: reports which concrete model Volcengine Ark Agent Plan's AUTO (ark-code-latest) is currently routed to, by sampling a 1-355 random-number distribution and matching it against bundled fingerprints of 16 candidate models.

↓ 0★ 0
Updated 10/23/2026
v0.1.2
dsh-plugin-pact

Plugin Pact

PACT governance plugin for DeepSeek Harness agent loops

↓ 0★ 0
Updated 10/23/2026
Av1.0.0
dsh-agents-init

Agents Init

Run an AGENTS.md build, repair, or audit entirely inside a subagent, so the main conversation never pays for the work.

↓ 0★ 0
Updated 10/23/2026
Pv0.13.0
dsh-personal-assistant

Personal Assistant

A proactive, human-like personal-assistant kernel for DeepSeek Harness: personal knowledge, stable and situational guidance, follow-through tasks, adaptive check-ins, resource discovery, and reusable experience.

↓ 0★ 0
Updated 10/23/2026
Wv0.1.6
@beforewave/dsh-with-chatgpt

With Chatgpt

Primary DSH entry for using ChatGPT with local code, Agent Helm engineering capabilities, and native DSH sessions.

↓ 0★ 0
Updated 10/23/2026
← Previous1…3536373839…52Next →
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in