Agent 連携 · DeepSeek Harness プラグイン — 139 ページ

Bridge DeepSeek Harness (dsh) into Feishu / Lark with streaming cards, project workspaces, approvals and scheduling

A DeepSeek Harness (dsh) plugin that bridges dsh agents to Feishu/Lark group chats — one group, one project directory.

Bidirectional Lark/Feishu control bridge for DeepSeek Harness

Lark/Feishu IM bot channel for DeepSeek Harness: chats drive agents, replies and approvals return as messages and cards

边做边学知识库：CRAG 式纠错检索（L1 Markdown wiki + L2 Hindsight）→ 未命中后台经 L3 限流联网补料 → 蒸馏落 staged → 两段式 commit 升入 L1。

Real-time code feedback for DeepSeek Harness — LSP, linters, formatters, type-checking, and structural analysis.

Visual lifecycle diagnostics for DeepSeek Harness plugins, Cordis Effects, and teardown outcomes

DSH Web GUI 轻量记忆系统：四个大写 Markdown 文件（USER.md / PROJECT.md / WORKLOG.md / CONVENTION.md）+ 两个动作（append 流水 / distill 沉降）。零外部依赖，只用 Node 内置模块；文件可视化、可 diff、可 git 管理。

Adobe Lightroom Classic for DeepSeek Harness (dsh) — mounts the lightroom-py MCP server as native dsh tools. Pure-config bundle inside the lightroom-py repo; the Python package comes from PyPI (pip install lightroom-py).

灵魂（Linghun）— 给 DeepSeek Harness 装一个会思考的自我。一张灵魂卡自定义四个字段：姓名、性格、沟通风格、其他——我是谁，由你亲手写；认知循环让判断有来处、反馈有归因、改进不断链；海马体把经历沉淀成知识与经验，踩过的坑下次不踩。不是人设卡，是判断力。

DeepSeek Harness 液态玻璃：点一下，整个界面通透起来。透明度随心调，背景图自由换——让你的 Agent 住进风景里。Liquid glass for DeepSeek Harness: one toggle to turn the whole shell translucent, with a live opacity slider and a custom full-page background.

Model-callable literature_search tool for DeepSeek Harness: wraps <工具目录>\paper-search.py (OpenAlex cited-sort + arXiv relevance) so the agent searches literature with one structured tool call instead of reading a skill body and running a shell command.

DeepSeek Harness frontend runtime verification: detect, run, preview, interact, diff, and prove local web apps.

DeepSeek Harness plugin: run subagents on a local llama.cpp model via the OpenAI Responses (/v1/responses) protocol, with a subagent delegation skill

Automatic provider/model fallback chains for DeepSeek Harness agents when LLM requests keep failing (retry exhausted, auth, quota, rate limit)

Nous Portal OpenAI-compatible LLM adapter bundle for DeepSeek Harness

Keep a DeepSeek Harness agent step alive through an exhausted account quota: after the base provider retry policy gives up on a QUOTA failure, retry the same request once per hour without an attempt limit, and publish the entering, leaving, and current state of that long-term retry to other plugins.

LLM API rate limiting, concurrency control, queuing, and adaptive cooldown for DeepSeek Harness

Local Markdown wiki as long-term memory for DeepSeek Harness — ported from llmwiki

DeepSeek Harness long-run mission suite: mission control, Long-Run Captain preset, patched LLM-as-a-Verifier, self-wake timer.

A readable isometric visual workspace for DeepSeek Harness agents

Always-on-top macOS desk pet for the DeepSeek Harness, with a native menu and DIY skins. Shows what your agent is doing: working, waiting, finished, failed. Seven skins, or make your own from one image. No dependencies.

Magic Context — DeepSeek Harness adapter (Host / Agent / Client + profile bundle). Loads the shared Magic SQLite store and registers ctx_* tools, /ctx-* commands, knowledge injection, historian/dreamer scheduling, compaction policy, and the Typert Remote surface for the DSH client.

智能体表情包系统：随插件内置默认表情包（零配置即用），也支持添加任意自定义表情包文件夹；智能体在回复中用 ::文件名:: 标记引用表情，Host 端自动转写为图片，让回复更生动可爱。