Integrations · DeepSeek Harness plugins — Page 88

GitHub Issue, Pull Request, and Actions MCP Bundle for DeepSeek Harness

钉钉 ↔ DeepSeek Harness(DSH) 双向通信桥接器：钉钉企业内部机器人(Stream 模式) + DSH /api 外部客户端协议

Command your DeepSeek Harness machines from your phone: approval push, remote dispatch, session board. Built on DeepSeek Harness.

Zero-knowledge WebSocket relay for dsh-dispatch: routes end-to-end encrypted frames between DeepSeek Harness (dsh) machines and phones, and fans out Web Push when the phone is offline.

Docker MCP Gateway 管理面板：经 SSH 在远端服务器执行 docker mcp 命令，管理 Docker MCP（列出/部署/移除 profile 中的 MCP、搜索/拉取 Docker MCP Catalog）。免构建（纯 ESM：Host 端 Node 原生 import + spawn python paramiko，Client 端零依赖纯 DOM）。

面向编码 agent 的版本化库文档检索：检索文档目录、选择版本、按主题提取带相关度排序的文档片段。以 MCP stdio server 形式运行，零运行时依赖，可被 dsh 等插件化 harness 加载

Zero-dependency MCP stdio server that generates real .docx files from a JSON spec — Chinese office-automation vertical for DeepSeek Harness.

Reconciliation: statements, invoices and ledgers that have to balance. Bank/credit statement PDFs and invoice batches into ledger-ready tables, then reconciled — on a shared key in integer cents, or with no shared key at all by amount, date window, reference numbers and fuzzy counterparty names.

Offline content-addressed proof for supplied DSH workflow duty separation

Record MCP once. Replay it safely inside DeepSeek Harness.

Save tokens on the DeepSeek Harness (dsh) — on-demand tools / MCP / skills loading, cache-stable, with a v3 retrieval matcher. Works as an agent preset (picker-selected 「省钱模式」) or as a dsh.bundle (global, installable via `dsh plugin add`).

Per-session Engram memory for DeepSeek Harness: spawns an engram MCP child per agent session rooted at the session workspace, registers mem_* tools per agent scope, and injects the Memory Protocol as a system-prompt section.

Environment facts for DeepSeek Harness — injects today, platform, os_version, and harness_version into system prompt variables

DSH 桌面端外部链接修复：点击 http/https 链接自动在系统默认浏览器打开。补上壳层缺失的 webview 链接点击接线，直连已修复的原生桥 shell.open-external（DSH 桌面端 5.3.6+，Windows 版）。

给纯文本模型装眼睛：把粘贴/拖入/本地图片交给外部视觉大模型，翻译成文字自动喂给主模型（DSH 官方 bundle 插件，图片进来文字出去）

DSH 客户端插件：登录你的法信(faxin.cn)账号，向 AI 提供结构化的法律条文 / 司法解释 / 裁判规则(审判意见) / 指导性案例检索工具

Cross-ecosystem aggregation base ('聚合的聚合'): syncs the GitHub dsh-plugin topic + npm registry into one open JSON index, queried by model tools, a CLI (dsh-feed), and a minimal stdio MCP server

DSH Feishu self-built app integration: receives text messages over the official Feishu WebSocket long connection, routes them to a DSH agent session, streams safe assistant text deltas through a Feishu CardKit card, and falls back to ordinary text when card streaming is unavailable. Acknowledges with an OK reaction (text fallback). Standalone Cordis bundle — no dsh source changes.

Feishu/Lark chat channel for DeepSeek Harness (dsh) — a sibling surface over dsh-base

DSH host plugin: a Feishu (Lark) bot bridge — send @dsh messages from Feishu on your phone or PC, and the local DSH agent runs them and replies back to Feishu

Talk to DeepSeek Harness from Feishu (飞书): each Feishu chat is bound to a series of DSH sessions, messages become turns, answers are pushed back, an interactive `/permission` card switches the session's file sandbox mode, a `feishu_ask` card asks the user a question with buttons plus a free-text note field (instead of the GUI dialog a Feishu user cannot see), `/model` sends a card whose buttons switch the session model, `/sessions` sends a card whose buttons switch back to an earlier conversation of the same chat (`/open <n>` still works), and `/new` / `/open` manage the chat's conversation generations. Permission requests are asked with a Feishu card too (/approval toggles the policy). Every outgoing message is a card, because Feishu plain text does not render Markdown. Inbound files and images are downloaded into the workspace. Zero runtime dependencies — hand-rolled protobuf framing over Node's built-in WebSocket. 在飞书里直接跟 DSH 对话。

DSH 飞书机器人桥接插件：通过飞书机器人与 DeepSeek Harness 对话——飞书消息进 DSH 会话由 AI 处理，回复以流式交互卡片实时发回（思维链/正文/工具链分区、打字机效果），无需公网 webhook

Feishu/Lark bridge plugin for the DeepSeek Harness (DSH): receive messages over WebSocket long-connection, run a dedicated agent to execute commands, and reply back to Feishu.

DeepSeek Harness 插件：与大肥鲸飞书机器人双向桥（阶段结论推送 / 飞书对话回流 / 结束总结）