港股资讯早报：交易日早盘前的聚源数据 Skill
介绍 SkillHub 技能「港股资讯早报」（@user_cc9425b3/jy-hk-stock-morning-brief-v2，v1.0.2，恒生聚源）。基于聚源数据 MCP 生成含恒指复盘、板块个股、公告、研报、南向资金、行业全景与外围市场的早报。说明 skillhub 安装命令、Cursor/Codex skills 目录、requires_api_key 注意点及典型提问示例。
Read Moredsh-mcp-client-manager：在 DSH 设置侧栏管理 MCP 客户端行
yiyunet 维护的 DSH 社区插件 @yiyunet/dsh-mcp-client-manager（1.0.4，MIT），在设置→侧栏提供「MCP 客户端」面板，管理 @deepseek-ai/dsh-mcp-client 下的 MCP 行：列出、添加、改写、启停、删除。受管行写入 $DSH_HOME/cordis.patch.yml 与 servers.json，点保存写盘、通常热生效；手写行走官方 configEditor/pluginManager。凭据不回传浏览器，零构建零运行依赖。安装：dsh plugin --profile web add @yiyunet/dsh-mcp-client-manager。
Read More开发者工具进行浏览器测试：用 Chrome DevTools MCP 在真实浏览器里验 UI
SkillHub 社区 Skill「开发者工具进行浏览器测试」（@user_e40a4360/bowser-testing-with-devtools，v1.0.0）指导 Agent 通过 Chrome DevTools MCP 在真实浏览器中检查 DOM、控制台、网络、性能与无障碍，并给出 UI/网络/性能调试流程与安全边界。本文介绍其定位、能力表、skillhub 安装命令（需 --dir 指向 Agent skills 目录）、MCP 配置示例及典型工作流，适用于前端与全栈在浏览器内的构建与排错。
Read Moredsh-plugin-weather：为 DSH 与 MCP 智能体提供全球天气与空气质量
dsh-plugin-weather 是 whaoran1018-cmd 维护的 DSH 联网工具插件，通过 get_weather 与 get_weather_forecast 提供实况、1–16 日预报、UV、日出日落及 PM2.5/AQI 等字段，数据主要来自 Open-Meteo，无需 API Key。同一核心可经 DSH bundle、MCP（bin/mcp-server.mjs）或 dsh-weather CLI 使用，支持 metric/imperial 与多语言摘要。推荐在 DSH 中用 plugin_manager 安装 github:whaoran1018-cmd/dsh-plugin-weather；使用前请审阅源码并以当前进程权限联网。
Read Moredsh-advancesearch：在 DSH 工作区里搜全会话，并跨 Agent 翻历史
dsh-advancesearch 是 wanetcn 维护的 DeepSeek Harness 工作流插件：在 Web 工作区侧边栏提供全局搜索，基于 sessionQuery 全文检索所有 DSH 会话标题与内容；可选扫描本机 Claude Code、Codex、ZCode 等 agent 会话目录。Host 暴露 GET /api/dsh-advancesearch，Client 提供搜索弹层；内置 MCP agent-session-search，供 agent 搜索、读取、打开应用或定位会话文件。安装需在 profile 目录 pnpm add 并在 cordis.patch.yml 启用。注意依赖 session-query 后端、插件以 DSH 进程权限读取本机数据，安装前请审阅源码与 MIT 许可证。
Read Moredsh-coderag：让 DSH Agent 按语义检索代码库
dsh-coderag 是 VermilionPasvikin 维护的 DSH 插件，以 MCP 提供 code_search、code_outline、code_index、index_status 四个只读工具。本地 SQLite FTS5 与 tree-sitter 分块索引，支持中文 bigram 与增量更新。安装需克隆 GitHub 仓库并执行 scripts/install.sh，再 dsh plugin add。默认不联网；安装前须知插件权限等同本机账号，DSH 文件权限档位不约束插件。
Read Moreget-arli-2：通过 Agent Skill 拉取阿里云相关搜索词
get-arli-2 是 SkillHub 社区技能 @user_201a3dfc/get-arli-2（v1.0.0，dev-programming），通过 SKILL.md 约定由 Agent 使用 MCP 拉取阿里云相关搜索列表。必填参数 robotId 对应 ALIYUN_RPA_RobotId，返回字符串列表。文章说明 SkillHub 元数据、skillhub install 安装到 Cursor 等 skills 目录的步骤，以及 MCP 与 RPA 配置需自行对接的限制。
Read MoreCoreness-DSH：为 DeepSeek Harness Profile 提供页面、日志与限流
Coreness-DSH（Vensus137）是 DeepSeek Harness 的工作流插件包，为单个 Profile 提供 MCP/Qoder/用量可视化页、变更日志、可选队列式限流，以及通过 prompt 注册表生效的树规则。安装为 plain files，无需构建；可通过 github:Vensus137/Coreness-DSH 或 dsh plugin add 安装，要求 dsh>=0.2.1-rc.2。适合集中管理 Profile 配置并保留审计记录的开发者。
Read Moredsh-mcp-manager：在 DSH 桌面版里图形化管理 MCP 服务器
DSH 官方 base bundle 自带 MCP 客户端，但服务器条目需手工写入 profile 的 cordis.patch.yml，解析失败会导致无法启动。dsh-mcp-manager 以独立 bundle 提供侧边栏「MCP 连接器」：列出、添加、删除、启用停用、测试连接与扫描导入；写入前用与 boot 相同的 js-yaml 校验，且只改标记区块。适用 Windows 官方桌面版，安装需在 desktop profile 执行 pnpm add 并登记 dsh.profile.bundles 后重启。MIT 0.1.0，维护者 VCPr0j3k7。
Read Moredsh-vedar：一条命令把堪察加旅游 MCP 接入 DeepSeek Harness
dsh-vedar 是 tourhabk-ui 维护的 DeepSeek Harness 联网插件，通过 cordis.patch 注册 @deepseek-ai/dsh-mcp-client，指向 vedarai.ru 公开 MCP。提供 20 个工具：线路与余位、火山与路线安全、天气、住宿装备交通、行程规划及人工确认的预订类请求。安装命令为 dsh plugin --profile web add github:tourhabk-ui/dsh-vedar，需 pnpm 与 Node>=18，重启 dsh web 后模型以 mcp__vedar__ 前缀调用。MIT 0.1.0，远端无认证、按 IP 限流，平台数据以俄语为主。
Read Moredsh-feyagate：在 DSH 里托管飞阳网关 MCP 引擎
dsh-feyagate 是 toddpan 维护的 DSH 工作流插件（npm 0.3.6，MIT）。它在宿主内托管 miloco-mcp-server：按清单下载校验、安装守护升级，经 MCP 门面将约 76 个家居工具接入对话，并提供服务、平台登录、授权与开放接口设置页。安装命令为 dsh plugin --profile <profile> add dsh-feyagate，随后在设置中安装后台服务并登录云平台即可在对话中控设备。插件开源，二进制运行时下载；使用前请审阅源码与 NOTICE。
Read More股市每日收评：基于聚源 MCP 的 A 股与港股收盘复盘 Skill
介绍 SkillHub 社区 Skill「股市每日收评」（@user_cc9425b3/jy-stock-daily-review，v1.0.2）。该技能通过 mcporter 连接恒生聚源 MCP，按交易日生成涵盖指数、涨跌统计、板块、资金流、公告舆情与技术面的 A 股与港股收评，支持 Markdown/HTML 双格式与数据溯源。文章说明 skillhub 安装路径、JY_API_KEY 与 MCP 配置步骤、典型自然语言问法及合规与交易日等使用注意。
Read Moredsh-skills-mcp-manager：在 DSH Web 设置页统一管理技能、MCP 与指令文件
SunlitCrack 维护的 DSH Web 插件，在设置页新增「技能与 MCP」入口，集中管理项目/用户技能目录、通过 dsh-mcp-client 真连启停的 MCP 服务器，以及 DSH 每轮读取的 AGENTS.md 家族指令文件席位。安装需 Node.js ≥22.19，推荐 dsh plugin --profile web add github:SunlitCrack/dsh-skills-mcp-manager 后 dsh web。适合不想分散改 JSON 与多路径技能的用户；MCP 凭证明文存 mcp.json，插件以 dsh 进程权限运行，使用前请审阅源码与许可证。
Read Moredsh-mcp-plugin：在 DSH Web 设置里管理 MCP 服务器
shinelon/dsh-mcp-plugin（@local/mcp-manager）是 DeepSeek Harness 的 Web 双面插件，在设置中新增「MCP 服务」页，对当前 profile 的 MCP 做列表、新增、编辑、删除与启停，展示 loader fiber 实时状态。变更写入 ~/.dsh/profiles/web/cordis.patch.yml 并由 HMR 热重载。支持 stdio 与 streamable-http 表单校验、CST 保注释与 !!js 透传、回环路由与原子写。安装示例为 dsh plugin --profile web add link:本地路径，需重启一次；MIT 许可，DSH >=0.1.7-rc.2。
Read Moredsh-windows-mcp-lite：用配置把 Windows 桌面自动化接到 DSH
dsh-windows-mcp-lite 是 sheilacraig 维护的 DSH 纯配置插件，通过 cordis.patch.yml 挂载内置 MCP 客户端，经 uvx 启动 windows-mcp@0.8.8，默认提供 16 个 mcp__windows__ 工具并排除五项与 DSH 重复的高权限能力。适用于 Windows 10/11 上的桌面自动化工作流；安装后需重启 DSH，建议先 npm run prewarm。无 Host 代码、无 HTTP 路由，适合重视审计的场景。
Read MoreMCP 极简管理：在 DSH Web 里一览并停用 MCP 服务器
dsh-mcp-mini-manager 在 DSH Web 设置侧栏提供与「插件」同级的 MCP 管理页，合并 Loader 与 cordis 补丁层列出全部 dsh-mcp-client 实例的状态、工具与简介，并通过 profile 层可恢复停用补丁逐个下线 MCP。文章基于官方 README 说明安装命令（GitHub / npm / 本地路径）、移除与恢复机制、安全边界及已知局限，便于智能体开发者在不手改 bundle 的前提下运维 MCP。
Read Moredsh-destinywind-mcp：DSH 下的 MCP 可视化管理与按需工具注入
dsh-destinywind-mcp 是 rickwindman 维护的 DSH Web 插件（admin-security），基于 dsh-mcp 1.12.0 修改。在 Settings → MCP 提供服务器增删改、启停、测试连接、工具勾选与 JSON 全量配置；支持 search/full 注入、环境变量与凭据托管、streamable-http OAuth，并只读展示 cordis.patch.yml 声明及接管/释放。安装通过 install_bundle 或 dsh plugin --profile web add link:<仓库路径>，需重启 dsh web 并硬刷新。插件以 DSH 进程权限运行，使用前请核对源码与 MIT 许可证。
Read Moredsh-project-config：用 .dsh/settings.yml 把子代理、命令、MCP 和 hooks 带进项目
DSH 默认可随项目读取 AGENTS.md 与 .dsh/skills，但子代理、斜杠命令、MCP 与 hooks 原先多在全局配置。RicardoYan 的 dsh-project-config（v0.2.1）从工作区 .dsh/settings.yml 与 settings.local.yml 加载上述能力，并提供 /project-config 诊断。命令与 MCP 仅在 dsh-workspace-trust 标记为已信任的工作区执行；适用于 DSH >=0.2.1-rc.1 且 <0.3.0。本文基于官方 README 说明安装、配置示例与安全边界。
Read Moredsh-plugin-relvato：在 DSH 里接入 Relvato 网站监控 MCP
dsh-plugin-relvato 是 Relvato 维护的 DSH bundle，仅通过 cordis.patch.yml 将托管 MCP（streamable-http，58 工具）挂入 profile，无本地构建与运行时代码。安装前需在 app.relvato.com 创建 API key 并设置 RELVATO_API_KEY，再用 dsh plugin --profile web add dsh-plugin-relvato 启用。agent 可添加站点、运行浏览器监控、读取 Core Web Vitals 与 fix brief，并受与仪表盘相同的权限与 destructive 标记约束。包版本 3.5.0，MIT 许可。
Read More基金对比分析：用 Agent Skill 生成七模块公募基金对比报告
「基金对比分析」是恒生聚源在 SkillHub 发布的 professional 类 Agent Skill（v1.0.1），基于恒生聚源 MCP 金融数据库，接收多只基金代码或简称，按分治原则逐标的调用工具，输出含基础信息、业绩、持仓风格、基金经理、费用、风险与投资建议的对比研究报告。需配置 API Key。本文说明定位、七模块能力、SkillHub CLI 安装命令及典型提问示例，并提示数据合规与耗时等注意事项。
Read Moredsh-control-center：在 Settings 里管理 MCP、Skill 与全局人设
dsh-control-center（MYanhui/DSH-plug-in）是 DeepSeek Harness 双面插件，在 Settings 提供 MCP 管理、Skill 管理与全局人设三页。MCP 通过 managed patch 块写入 Loader 行；Skill 编辑 dsh-skill-filesystem 扫描的全局根目录并支持从 GitHub 归档拉取；人设以 Markdown 库存储并注册 system-prompt section。安装可用 `dsh plugin --profile <name> add dsh-control-center` 或 Windows 脚本/手动 mount。适用需可视化维护全局资源且接受插件以 dsh 进程权限运行的开发者。
Read MoreThesis Tutor MCP：把论文写作辅导挂进 MCP 客户端
Thesis Tutor MCP 是 SkillHub 上的社区教育类 Skill，基于 MCP 与 FastMCP，向兼容客户端暴露 9 个论文辅导工具：选题、大纲、章节批注、解读反馈、交稿检查、答辩模拟及参考文献查询。无需 API Key，可通过 skillhub CLI 安装到 Cursor、Claude Code 等 skills 目录。适合将选题到答辩的准备步骤工具化，最终以学校与导师要求为准。
Read Moredsh-plugin-yuantoutao：在 DSH 里接淘宝、京东、拼多多的对话购物 MCP
MeshTok 维护的 dsh-plugin-yuantoutao（源头淘）为 DeepSeek Harness 提供电商 MCP：自然语言搜货、可解释比价、本地精选货架与跨淘系/京东/拼多多工具，可选联盟转链与履约闭环。只读购物可无 API key 使用；需实时扇出、下单与分成时再配置 YUANTOUTAO_API_KEY。安装：dsh plugin add github:MeshTok/dsh-plugin-yuantoutao，重启 dsh web 或 headless。MIT，dsh>=0.2.0。
Read More民事上诉状（AI结构化书写）Plus：按模版生成上诉状并导出 docx
SkillHub 社区 Skill「民事上诉状（AI结构化书写）Plus」（v1.0.2）基于一审裁判文书与用户材料，按模版结构化书写《民事上诉状》，支持对话修改与导出 docx。引用法律法规时依赖「华宇元典法律数据」MCP，需在 WorkBuddy 连接器中手动连接。安装坐标为 @user_08c66eae/civil-appeal-petition-ai-structured-writing-plus，按 skillhub.cn/install/skillhub.md 配置；标签 requires_api_key=false。
Read Morelimonene-mcp：在 DSH 里只读接入 Amazon 卖家数据
limonene-mcp（limonene-dsh）是 M-Abozaid 维护的 DeepSeek Harness 社区包，通过浏览器 OAuth 连接 Limonene 托管 MCP 端点 https://limonene.app/api/mcp，在 DSH Web 试点中提供六项只读工具：店铺概览、告警、Buy Box、FBA 库存、销售趋势与收入费用分解。安装命令为 dsh plugin --profile web add github:M-Abozaid/limonene-mcp 后执行 dsh web，需 Node 22+ 与已同步的 Limonene 店铺。文档要求选择 Read only，勿在 DSH 试点启用 Read and manage。插件以 DSH 进程权限运行，MIT 仅覆盖连接器包文件。
Read More基金经理画像：用恒生聚源数据生成综合画像报告
基金经理画像是恒生聚源在 SkillHub 发布的企业级 Skill（v1.0.2），基于恒生聚源 MCP 金融数据库，从投资理念、管理业绩、资产配置风格、选股、择时与风险控制等维度剖析基金经理，输出《基金经理综合画像报告》。用户以基金经理姓名或基金代码触发即可。安装需按 skillhub.md 安装 @user_cc9425b3/jy-fund-manager-profile，并配置 requires_api_key 所要求的 API 密钥。适合基金研究与投顾场景下的结构化评估草稿，不能替代独立尽调。
Read MoreDSH Rice on a Stick：用 MCP 操作 DeepSeek Harness 真实会话
DSH Rice on a Stick（dsh-rice-on-a-stick）是 Leawind 维护的 MIT 工作流插件，在 dsh >= 0.2.1-rc.2 上暴露 Streamable HTTP MCP（默认 127.0.0.1:8090/mcp），经 sessionController 创建与编排与会话 UI 一致的真实 session，支持 session_submit/poll/result 异步任务、审批策略与 rice_ 自举工具。本文基于官方 README 说明安装 link、客户端连接、典型工具链路与安全配置，并提醒本地代码执行风险与权限边界。
Read Moredsh-skills-mcp-panel：在 Web GUI 里统一管理 Skills 与 MCP
dsh-skills-mcp-panel 是 KyattoCat 维护的 DSH Web 客户端插件（v0.2.0，MIT）。在 Settings → Agent presets → Skills/MCP 中按目录分组列出部署内全部 Skill 与 MCP，支持搜索、状态灯与热切换：Skill 通过 SKILL.md frontmatter 四态控制，MCP 在 cordis.patch.yml 标记块写入 disabled，live profile 下同会话生效。新装需重启 dsh web 一次。安装命令见官方 README 的 dsh plugin --profile web add。
Read Moredsh-skill-mcp-panel（jipika 设置分栏改版）：在设置窗口管理 Skill 与 MCP
jipika/dsh-skill-mcp-panel 是基于 Fishquito7/dsh-skill-mcp-panel 2.1.0 的本地改版：将 Skill 与 MCP 管理从 Web 侧栏+中央主区改为 settings.section 设置分栏，并去掉 layout 硬依赖。文章说明与上游差异、Skill/MCP 能力与 dsh-panel CLI、上游 README 安装命令的对照，以及同 entry id 不可并存等注意事项。
Read Moredsh-mcp-hub：DSH 下的 MCP 一键部署与工具中枢
dsh-mcp-hub 是 JD579g 维护的 DSH 工作流插件，在 Settings → MCP Tools 中一键安装 MCP、按需搜索官方 Registry/npm/GitHub，连接运行中的 Harness 无需重启。提供 /mcp doctor、四级权限、与 DSH 原生工具去重，以及 7 个零依赖内置 MCP 服务（61 工具）；Windows 优先处理 npx.cmd 与 pwsh 等差异。安装：dsh plugin --profile web add dsh-mcp-hub 或 install.ps1/install.sh，需 Node >= 20，MIT 1.2.0。
Read Moredsh-onto-space-panel：在 DeepSeek Harness 里选 Onto 空间并管理对话上下文
hope4b 维护的 DSH 插件 dsh-onto-space-panel（v1.2.1）在 Web/Desktop Harness 输入框上方提供 Onto 空间选择器与侧栏，可挂空间声明、resident agent 章程、记忆工件及对象到当前聊天，上下文按会话与 realm 隔离。支持首次 MCP 连接向导或手动 cordis.patch.yml 配置 serverName onto。安装 release tgz 或 github 标签后需重启 Harness；集成针对 Harness 0.2.0-rc.2 验证。界面为俄语，依赖 Onto MCP 多项 list/search/get 工具。
Read MoreMini-HIDS：用 Agent Skill 把单机防火墙与入侵检测接到对话里
Mini-HIDS 是 SkillHub 社区 Skill（@user_9a8108cb/minihids，v1.2.6，it-ops-security），将轻量 Linux 主机检测与防火墙封禁封装为 Agent 可调用的 MCP 能力。文章基于官方页面与 Skill 包内 SKILL.md，说明 SSH/Web 检测、SQLite 黑名单、六项 MCP 工具，以及 skillhub 安装、config.json 配置与 daemon/MCP 启动步骤，并区分适用场景与安全注意点。
Read MoreOlaChill DSH 插件：在 Harness 里查价与比价日本旅行服务
olachill-dsh-plugin 是仅含配置的 DeepSeek Harness 插件，通过内置 MCP 客户端连接 OlaChill 公共服务端，无需 API Key 即可在对话中搜索日本游、票券、旅馆、接送、包车、直升机、golf 与 eSIM 等 13 类工具。本文说明定位、工具列表、官方安装命令与示例 prompt，并提醒只读与询价确认、起价说明及安装前审阅源码与权限。
Read Moredsh-computer-use-linux：在 DSH 里用原生工具控制 Linux 桌面
社区插件 @brucezhang1993/dsh-computer-use-linux 通过内置 dsh-mcp-client 将 computer-use-linux MCP 接入 DeepSeek Harness，暴露 18 个 mcp__cul__* 工具，支持 AT-SPI、截图、窗口管理与 Wayland/X11 输入。文章说明架构、安装（GitHub Packages 与 dsh plugin 命令）、doctor 验证及 compositor 相关注意事项，适合在 Linux 工作站上做 GUI 自动化的 DSH 开发者。
Read Moredsh-codex-chrome：在 DSH 里复用 Codex 原生 Chrome 控制能力
dsh-codex-chrome 是 BruceZhang1993 维护的 DeepSeek Harness bundle 插件，将 OpenAI Codex 的 chrome@openai-bundled 运行时迁入 DSH，通过 bin/chrome-mcp.mjs 注入 turn 元数据并暴露十一个 mcp__chrome__* 工具与 control-chrome skill。智能体可操作用户本机 Chrome 的标签、登录态、无障碍树、截图与输入，工作流为 observe→act→observe。CLI 安装：dsh plugin --profile web add 绝对路径；需 Linux、Codex 缓存、浏览器扩展与 native host。MIT，0.1.0，Node≥20。
Read Moredsh-skill-mcp-tool：在 DSH 里统一管理 Skills 与 MCP
dsh-skill-mcp-tool 是 artorias-zj 维护的 DeepSeek Harness 插件（v1.0.2，MIT），在设置与侧栏双入口提供同一管理面板，覆盖四个标准 Skills 根目录与 MCP 客户端行。支持通过文件改名热加载启用/停用技能、Git 浅克隆安装与更新（可选命令级 Git 代理）、cordis.patch.yml 受管块内 MCP 热挂载，以及配方/完整两种备份导出。安装需 dsh plugin add 官方命令并在启用后重启 DSH；插件以进程权限读写配置并调用 git，使用前请审阅源码。
Read Moredsh-onedev：让 DSH 通过 MCP 直连 OneDev 管理 REST API
dsh-onedev 是 Altermoe 维护的 DeepSeek Harness 插件，内置 dsh-onedev-mcp，经 dsh-mcp-client 将 OneDev 管理 REST API 暴露为 mcp__onedev__ 工具（61 个，含多环境、文档查询与 onedev_api_request）。支持 Token 或账号密码、只读模式，bundle 安装后可在 Settings→OneDev 配置凭据。需 Node ^22.19.0 与 DSH ^0.2.0-rc.2。本文基于 GitHub README 与 dsh-integration 文档整理安装与典型调用。
Read Moredsh-as-mcp：把 DeepSeek Harness 暴露为 MCP 服务
dsh-as-mcp 由 xiseliuli 维护，在 DSH 宿主进程内提供 MCP 端点，使 Claude Code、Codex、脚本或 CI 可通过标准 MCP 创建工作区与会话、下发编码任务、读写文件并执行 shell。会话与 UI 中一致，走同一沙箱与权限策略。本文基于官方 README 介绍安装命令、HTTP/stdio 接入、工具表与安全边界，并给出典型 workspace → session → session_prompt 流程。
Read More数电发票开票 Skill：用自然语言走完企享云数电开票流程
介绍 SkillHub 社区 Skill @user_bbae3f97/qixiangyun-invoice-issuance-skill（v1.0.14）：用自然语言将开票意图解析为结构化草稿，衔接税局登录、云抬头、智能赋码、确认开票，并通过企享云 MCP 完成蓝字发票及人脸认证、已票查询与红冲。说明 skillhub CLI 安装命令、各 Agent skills 目录与 requires_api_key 配置要点，并给出官方示例话术。
Read More基金经理观点与持仓一致性分析：用聚源数据检验「说法」与持仓是否一致
SkillHub 企业级 Skill「基金经理观点与持仓一致性分析」（@user_cc9425b3/jy-fund-consistency-analysis-v2，v1.0.1，恒生聚源）基于聚源金融库与 MCP API，从定期报告提取投资观点，对照基金实际持仓，在行业配置、投资风格、选股逻辑、风险控制等维度做一致性评分并生成标准化报告。需 API Key。安装见 skillhub.cn/install/skillhub.md，用 skillhub install 指定 Cursor 等 skills 目录。适用于言行一致性、可信度与匹配度分析。
Read Moredsh-rea-mcp：用配置把 REA 逆向分析挂进 DeepSeek Harness
dsh-rea-mcp 由 sheilacraig 维护，是仅含 cordis.patch.yml 的 DeepSeek Harness 工作流插件，将内置 MCP 客户端通过 stdio 指向 rea-agents@6.4.0。无 Host 代码与 HTTP 路由，适合在 DSH 智能体中调用 mcp__rea__* 工具做本地二进制、JS、.NET、APK 等分析。需 DSH >=0.2.0-rc.1 与 Node 22+/24+/26+，安装后建议 npm run prewarm 并重启。分析在你本机与用户权限下执行，安装前应阅读 patch 与许可证。
Read More行业速报：基于聚源库生成可溯源的行业 24 小时速报
行业速报是恒生聚源在 SkillHub 发布的 professional 类 Skill，通过聚源金融数据库与 MCP 为指定行业生成近 24 小时速报，默认 Markdown、可选 HTML。输出覆盖备选简读、重要新闻、企业动态与投融资四块，信息可溯源并带原始发布时间。本文说明其定位、核心能力，以及按 skillhub install 安装到 Cursor 等目录的步骤与官方示例问法；标签 requires_api_key 表示需先配置 API 与 MCP。
Read Moreconsole-translator：把 Ardot 里的中文控制台文案翻成可上线的英文
console-translator 是 SkillHub 社区 Skill（@user_5d52b855/console-translator，v1.0.0），面向 Ardot 选中范围内的中文 UI 文案，产出控制台、仪表盘与 SaaS 风格英文。它做页面级行业推断，优先 references/dict.md，并对照推断行业内海外前三产品用语，再经本地化与精简改写后通过 Ardot MCP 写回设计稿。安装需 skillhub CLI 与 --dir 指向 Agent skills 目录；标签 requires_api_key=false。适合 Ardot 协作且需术语统一的 B 端界面团队。
Read More电脑全自动控制：用 Rebind MCP 让 Agent 操作本机桌面
SkillHub 技能「电脑全自动控制」（@user_6be3c359/rebind-pc-auto-control，v1.0.0）基于 Rebind 可编程输入层与 @rebind.gg/mcp-server，面向 Windows/macOS/Linux。当 Agent 需要看屏幕、移动鼠标、点击、输入、滚动并操作窗口，以完成自动填表、批量处理文件、无 API 的 GUI 或跨应用流转时，该 Skill 提供 MCP 接线、环境自检与安全相关的全自动工作流说明，无需 OpenAI Codex 账号、本机运行。安装可遵循 skillhub.cn/install/skillhub.md，使用 skillhub install 并指定 Agent 的 skills 目录。
Read More股票市场热点解读：用聚源数据做 A 股热点与板块分析
恒生聚源在 SkillHub 发布的「股票市场热点解读」（v1.0.1）通过 MCP 连接聚源金融数据库，覆盖热点识别、深度解读、产业链映射、资金与情绪追踪、操作建议及风险预警七类能力。本文说明其定位、核心模块、skillhub 安装命令与 Cursor 等客户端 skills 目录，并给出官方问题示例。Skill 标注 requires_api_key=true，使用前需配置聚源相关 API。
Read MoreDEEPQ-ETF分析：用 Agent Skill 串起单只 ETF 的多维调研
DEEPQ-ETF分析是 SkillHub 专业类技能，面向指定 ETF 做基本信息、实时行情、技术面、基本面、业绩、底层资产、消息面与异动归因分析。用户询问 ETF 概况、是否值得买入或大涨大跌原因时触发，通过 guessFundCode 及 etfBasicInfo、etfLatestPrice 等 DeepQ MCP 工具拉数。文章说明技能定位、工具矩阵、与同系列 DEEPQ 技能一致的 MCP 配置思路，以及典型自然语言用法与 API Key、合规等注意事项。官方入口：https://api.skillhub.cn/user_e7dd0698/etf-analysis。
Read More股票对比分析：用恒生聚源数据生成 A 股多标的对比报告
SkillHub 企业 Skill「股票对比分析」（v1.0.2，恒生聚源）在用户需要对比 2–5 只 A 股时触发，经恒生聚源 MCP 金融数据库拉数，输出 PDF 与 Markdown 双格式报告。章节覆盖公司简介、基本面、技术面、资金面、消息面，并强调数据可溯源与日期标注。安装坐标为 @user_cc9425b3/jy-stock-comparative-analysis，需按 skillhub.cn 安装文档配置，标签要求 API Key。
Read MoreElastic / Kibana MCP 服务器 Skill：把 Kibana Agent Builder 接到 MCP 宿主
SkillHub 社区 Skill @user_3c6cb52e/elastic-kibana-mcp-server（v1.0.0，it-ops-security）引导 Agent 按 Elastic 官方文档配置 Kibana Agent Builder MCP：端点 /api/agent_builder/mcp、Space 路径、API Key 与 OAuth 鉴权，并在 Cursor 等宿主用 mcp-remote 连接。Skill 不替代 Kibana 服务，安装用 skillhub install 与 --dir；敏感环境需先核对上游 elasticsearch 与 Elastic 文档。
Read MoreControl Unity Editor scenes and assets through MCP for Unity：用 MCP 在编辑器里动手
介绍 SkillHub 社区 Skill「Control Unity Editor scenes and assets through MCP for Unity」（v1.0.0，dev-programming）。它指导 MCP 兼容 Agent 在装好 CoplayDev MCP for Unity 后检查场景、管理资源、编辑脚本与运行测试。文章说明 Skill 与上游工具链的分工、先决条件（Unity、Python 3.10+、MCP 客户端）、Unity 包 Git URL 与 Window 菜单配置、以及 skillhub install 到 Cursor 等 skills 目录的命令，并给出上游 Quickstart 示例与文档链接。
Read More飞书项目 MCP 对接：用本地业务上下文减少查字段
飞书项目 MCP 已能连上时，建单改单仍常卡在字段、枚举和角色对齐。SkillHub 社区 Skill「飞书项目 MCP 对接」（@user_acfc9128/feishu-project-mcp-playbook，v1.0.0）在 feishu-project MCP 连接完成后，引导 Agent 读取本地业务上下文，复用字段映射、角色、默认值与常用候选项。文章说明前置依赖、SkillHub 安装路径（含 Cursor/Claude Code/Codex 目录示例）及持续维护上下文的做法，适合工单密集、空间配置相对稳定的团队。
Read More会议纪要助手：腾讯会议 MCP 上的五段式自动闭环
会议纪要助手是 SkillHub 社区 Skill，在腾讯会议 MCP（预约、录制、转写、智能纪要等）之上提供会前到会中、会后、洞察、场景五段式 AI 工作流。v2.7.0 起支持 auto 编排：输入会议号或链接可自动完成纪要、待办、质检、脱敏、导出、腾讯文档同步与跟进邮件草稿，并含断点续跑与失败降级。另含数据驱动结构库、BANT/MEDDICC 模板、跨会议问答与健康度趋势。安装需 skillhub CLI 指向 Agent skills 目录，并配置 requires_api_key。
Read More全球法律检索·小律同学 AI：面向出海与国内合规的 Agent Skill
介绍 SkillHub 企业技能 @org-tz5v519f/mcp-law-search-skill（v1.0.5）：小律同学发布的法律检索与文书生成 Skill，覆盖国内法律法规、司法解释、指导案例、量刑标准，以及 168 个国家/地区的条约、外国法与跨境合规要求，并支持 697 类文书起草。说明 skillhub CLI 安装、Cursor/Claude Code/Codex 的 skills 目录、OAuth 或 API Key 鉴权，以及中文关键词检索习惯与不适用诉讼策略等限制。
Read More基金分析报告：用 Agent Skill 生成可溯源的基金 Markdown 研报
基金分析报告是恒生聚源在 SkillHub 发布的企业级 Skill（v1.0.1），基于 jy-financedata-api MCP 生成 Markdown 基金深度研报，覆盖概况、业绩与同类排名、资产配置、基金经理、基金公司、投资建议六模块，数据可溯源并带时间戳。需 API Key，可通过 skillhub install 安装到 Cursor 等 Agent。适用于基金分析、持仓与风险评估、同类排名等对话场景。
Read MoreTencent Cloud APM：在 Agent 里用 MCP 做调用链与性能分析
Tencent Cloud APM（@tencent-adm/apm-performance-analysis，v1.0.3）由腾讯云 APM 团队发布，通过 ApmClient.SendMCPMessage 连接 APM MCP Server，配合内嵌 SDK 与 apm_mcp_client.py 完成 ping、列工具与 call-tool。文章说明 Skill 定位、核心能力、SkillHub 安装路径、TENCENTCLOUD_SECRET_ID/KEY 配置，以及 list-tools 与 DescribeApmOverview 示例调用流程，并提示 requires_api_key 与交互确认规范。
Read MoreBGPT 论文检索：用 MCP 拿全文结构化实验数据
BGPT 论文检索是 SkillHub 上的 knowledge-management 类 Skill（v1.0.0，智雨科技发布），引导配置 BGPT 远程 MCP，调用 search_papers 检索科学论文并返回 25 个以上结构化字段（方法、结果、样本量、质量评分、结论等）。文章说明 skillhub 安装路径、Claude 的 MCP 配置、典型查询示例，以及免费额度与付费规则，适合文献综述与证据表场景。
Read More宏观经济分析：用 Agent Skill 生成中国宏观月度/季度深度报告
恒生聚源在 SkillHub 发布的「宏观经济分析」技能（@user_cc9425b3/jy-china-macro-analysis，v1.0.2）面向完整中国宏观月度/季度深度报告。基于恒生聚源 MCP 完成数据获取、图表生成、逻辑分析与合规输出。适合明确要写整份报告的场景；单项 CPI、PMI 等解读不必触发。安装需 skillhub CLI、--dir 指向 Agent skills 目录，并配置 requires_api_key 所要求的 API Key。
Read More万得金融数据：让 Agent 通过 Wind MCP 查行情与基本面
万得金融数据（@user_db7424f5/udb7-wind-financial-data，v1.0.0）是 SkillHub professional 类社区 Skill，通过本地 cli.mjs 调用 Wind 七个 server_type，覆盖 A/港/美股、基金 ETF、指数板块、债券、公告新闻与宏观 EDB；不含欧股、日股与加密。文章说明 SkillHub CLI 安装路径、WIND_API_KEY 配置，以及 search_stocks、get_stock_price_indicators 等可复现 CLI 示例与多标的、日期字段等注意事项。
Read Moreifind-finance-data2：让 Agent 按自然语言查同花顺 iFinD 数据
ifind-finance-data2 是 SkillHub 社区专业类 Skill，封装同花顺 iFinD 金融数据 MCP，支持股票、基金、宏观 EDB、新闻公告、债券、港美股与指数板块查询及智能选股选基。文章说明其定位、按 server_type 划分的核心能力、通过 skillhub.md 安装坐标 @user_c163d6d4/ifind-finance-data2、mcp_config.json 配置 auth_token，以及 Node.js/Python 调用示例与并发限制，便于在 Cursor 等 Agent 中落地投研取数。
Read More微云 MCP：让 Agent 通过 MCP 操作微云网盘
微云 MCP（@tencent-adm/tencent-drive-mcp）是腾讯微云在 SkillHub 发布的 dev-programming 技能，指导 Agent 经 MCP 调用 weiyun.list、upload、download、gen_share_link 等 14 个工具，覆盖列表、搜索、移动、分享与 FTN 分块上传。安装使用 skillhub install 并配置 mcporter 与 mcp_token；附带 setup.sh 授权与 upload_to_weiyun.py。仅适用于 weiyun.com 网盘，不涵盖腾讯文档与其他云盘。
Read More金蝶云星空 MCP 连接器：在 WorkBuddy 中接入第三方 kingdee-mcp
金蝶云星空 MCP 连接器是 SkillHub 社区 Skill（@user_4643e8be/kingdee-mcp-connector，v1.0.2），用于在 WorkBuddy 中连接并配置基于第三方开源 kingdee-mcp 的金蝶云星空 MCP Server，非金蝶官方出品。接入后可用自然语言查询或操作采购、销售、库存、生产、成本、固定资产等 ERP 单据。安装可参考 skillhub.md，使用 skillhub install 并指定 skills 目录。标签 requires_api_key=true，凭据需按技能文档安全配置。
Read More钉钉文档skill：在 Agent 里操作钉钉知识库与文档
钉钉文档skill（@user_924c5114/dingtalk-docs-skill，v1.0.4）是 SkillHub 社区办公类 Skill，需 API Key，支持 Markdown 推送、拉取与覆盖/追加更新、块级编辑、搜索列出、下载附件、导出 PDF/Word、节点权限及文档夹管理，并可初始化钉钉文档 MCP。本文说明定位、能力与 SkillHub CLI 安装路径，并区分与飞书/语雀及钉钉 IM 的边界。
Read More数据库操作 Skill：把 MCP 数据库工具接进 Agent 工作流
SkillHub 社区 Skill「数据库操作」（@user_16c0d092/database-mcp v1.0.1）是一份 SKILL.md 指南，教 Agent 使用 @turingdance/database-mcp：connect_db、list_tables、describe_table、query 与 execute 五个工具，支持多种 SQL 库与环境变量配置。文章说明 SkillHub 与 npx 安装方式、MCP 配置示例、先测连再查询的工作流，以及写操作 confirm 与风险分级。适合在 Cursor 等环境规范数据库协作的开发者；无需 API Key。
Read More政策解读：用恒生聚源 MCP 做宏观与监管政策分析
政策解读是 SkillHub 上由恒生聚源发布的企业级技能（v1.0.2），基于恒生聚源 MCP 金融数据库，用于理解宏观、产业与监管政策，分析对金融市场的影响并支持查询政策原文。需配置 API Key。可通过 skillhub install 安装到 Cursor 等客户端 skills 目录，或用官方提示词让 AI 按 skillhub.md 安装。适合研究员与需要政策—市场联动分析的场景。
Read More通达信股票数据：在 AI 对话里查行情、研报与选股
介绍 SkillHub 社区 Skill「通达信股票数据」（@user_d8831e99/tongdaxin-mcp，v1.0.4）。它是通达信连接器，支持全球股票行情、研究报告、公告资讯与宏观信息，以及通达信指标、条件选股、基本面分析、同行业对比和智能选股筛查。文章说明 Skill 定位、核心能力、SkillHub 官方安装提示词与 API Key 要求，并给出对话式典型用法与合规注意。
Read More恒生聚源金融数据：用中文自然语言驱动 GilData 研究链路
恒生聚源发布的 SkillHub 专业类 Skill（v1.6.0），在 GilData MCP 十一工具之上提供路由、消歧、组合调用与输出整合，覆盖 A 股港股基金行情财务、公告研报、宏观行业、资讯舆情、会议纪要与智能筛选等中文研究场景。本文说明定位、工具映射、skillhub 安装路径与典型提问方式，并注明 requires_api_key=false 及使用注意。
Read MoreMcporter Skill：用 CLI 管理并调用 MCP 服务器与工具
Mcporter 是 SkillHub（@clawhub_steipete/mcporter）上的开发类 Skill，源自 clawhub 发布者 steipete，教 Agent 使用 mcporter 命令行直接管理 MCP：list/schema、HTTP 与 stdio 调用、OAuth 与 config、daemon，以及 generate-cli 与 emit-ts。默认配置为 ./config/mcporter.json，机器可读输出可用 --output json。适合终端调试、CI 与胶水代码生成；需本机已安装 mcporter CLI。
Read Moredsh-mcp-unified-panel：DeepSeek Harness 能力库
`dsh-mcp-unified-panel` 是 DeepSeek Harness 生态中的一款 All-in-One 插件，旨在解决 MCP 服务器与技能管理的现有障碍。它具备原生 OAuth 2.1 客户端驱动，通过特征识别替代包名白名单，确保官方与第三方行同屏可见且安全隔离。该插件提供精简的界面交互与 CLI 工具，支持 MCP 连接的一键新增、连通性测试及技能的浏览与启用，实现了运维、支持与管理的统一，无需依赖外部配置。
Read Morefigma转高保真html：用 Agent Skill 把 Figma 设计稿落成响应式 HTML
figma转高保真html（@user_24c316b5/figma-to-html，v1.0.0）是 SkillHub 社区 dev-programming 类 Skill，通过 Figma MCP 读取设计数据与图片，并结合参考截图生成高保真响应式 HTML。支持「figma 转 html」等关键词或 Figma 链接触发。安装可参考 skillhub.cn/install/skillhub.md，使用 skillhub install 并指定 Cursor 等客户端的 skills 目录。需预先配置 Figma MCP；元数据标注 requires_api_key=false。
Read MoreCodex DSH Bridge：通过 MCP 将 Codex 会话连接到 DeepSeek Harness Web
journ10/codex-dsh-bridge 是一个 HTTP MCP 服务器，旨在连接本地 CLI 工具（如 Codex）与 DeepSeek Harness (DSH) Web 界面。它允许通过 MCP 协议管理 DSH 会话的生命周期，包括新建任务、发送消息、等待结果及查询状态。 安装需 DSH 版本 0.2.1-alpha.2，使用 `dsh plugin` 添加，默认监听 3080 端口。配置时需在 Codex 中指定 URL 并设置环境变量获取令牌。该插件支持同步和异步模式，但不支持非 Web profile。
Read MoreWorkBuddy 新手指南：入门文档与三套可运行脚本
SkillHub 社区 Skill「WorkBuddy最快上手指南」（@user_57af6244/workbuddy-newbie-guide，v2.0.2）面向刚接触 WorkBuddy 的开发者，提供界面地图、四种能力扩展说明、30+ 指令模板及省积分避坑，并附带环境体检、一键立规矩（项目规范与记忆骨架）、MCP 连接器配置（4 个常用模板、写前自动备份）三套可运行脚本。requires_api_key=false。安装时按 SkillHub 提示引用 skillhub.md 安装上述坐标。适合需要系统化入门与项目初始化流程的用户。
Read More法律智能检索 Skill：标准 RAG 策略、流程与输出规范
SkillHub 社区 Skill「法律智能检索-策略&流程&工具映射&输出规范」（@user_c715c6d2/liuye-legal-intelligent-search，v1.5.2）将法律事务处理整理为标准 RAG：并行检索北大法宝与 IMA 知识库，流程为问题识别、检索、Legal 分析、按规范输出，并含北大法宝 MCP 工具映射。输出须含法律依据、分析、风险评估、建议与免责，适用于法规查询、案例分析、合同审查与风险分析。安装需 skillhub CLI 与 --dir 指向 Agent skills 目录，标签 requires_api_key=true。
Read More基金投资组合配置+顾问：本地运行的基金智能顾问 Skill
介绍 SkillHub 社区 Skill @user_9d5a2a39/fundadvisor（v12.0，包版本 1.0.47）：纯本地零配置的基金组合录入、X 光透视、优化与回测，可选 DeepSeek 观点蒸馏；v12.0 新增外部 MCP（mcp_sources.json 与 FUND_ADVISOR_EXTERNAL_MCP_*）、类别路由与净值缓存溯源，未配置时静默降级。含 SkillHub CLI 安装路径、Cursor/Claude/Codex skills 目录及典型对话用法。
Read More企微文档读写全能小助手：用 MCP 读写 doc.weixin.qq.com
介绍 SkillHub 社区技能 @user_5f5835e5/wecom-doc-skill（v1.0.17）：内置企业微信文档 MCP 服务与 setup.sh，无需 git clone。说明 Node 18+、Playwright 扫码登录，以及 wecom_doc_fetch/write 等工具对 doc、sheet、smartsheet、slide、mindmap、flowchart 的读写能力；给出 skillhub install、~/wecom-doc-mcp 安装与 Cursor 等 MCP 配置路径及典型用法。
Read More微信公众号终极工作台（全流程一站式）：把公众号创作串成一条 Agent 工作流
SkillHub 社区 Skill「微信公众号终极工作台（全流程一站式）」（@user_f1de10d1/wechat-mp-suite，v1.0.2）将搜索、爬虫、写作（刘润风格/爆款推文/真人写作）、洗稿、AI 配图、排版与本地/远程 MCP 发布整合为一条 Agent 工作流。本文基于 SkillHub 公开简介与安装文档，说明定位、八大模块、skillhub CLI 安装路径（含 Cursor 等 skills 目录）及典型触发方式，并提示 requires_api_key 与合规使用注意。
Read More装前查 · Skill安全风险速查：安装中文 Agent 工具前先查风险
装前查 · Skill安全风险速查（@user_34343d1f/zhuangqiancha，v1.0.3）面向 Skill、MCP、插件等中文 Agent 工具，基于 zhuangqiancha.com 对公开代码与声明的评估，从权限透明、隐私去向、作者信誉、安全档位四维度回答「是否安全」「作者高风险工具」等问题。零 API Key，纯文本回复。可通过 skillhub 安装到 Cursor 等 skills 目录，适合安装前快速核对。
Read More微信小程序AI开发模式生成Skills：从源码生成 wx.modelContext 技能分包
SkillHub 技能 @tencent-adm/wxa-skills-generate（v0.2.3，微信小程序团队）用于静态分析小程序源码（含混淆），识别业务步骤并提取网络接口与 JSAPI，生成符合 wx.modelContext 的 skills/ 分包（原子接口与组件），并集成 app.json、project.config.json。适用于页面能力原子化、派生 MCP 工具等场景；生成后须交棒 wxa-skills-validate 校验。无需 API Key，可通过 skillhub install 安装到 Cursor 等 Agent 的 skills 目录。
Read MoreQuantitative Analysis Tool for Stocks, Funds, and Futures: Full-Asset Quant and Multi-Source Investment Research Skill
Introduces the SkillHub community skill @user_9d5a2a39/aistockresearcher (package version 1.0.45, with embedded v11.0). It covers quant analysis of stocks, funds, futures, and indices across 12 global markets: multi-source free data from Tencent, East Money, Sina, Xueqiu, Tonghuashun, yfinance, and others, with cross-validation and anomaly detection; 11 AI interpretation templates, multi-timeframe forecasting, and 20 MCP tools, with optional external MCP environment-variable bridging; plus v9.5 Expert Roundtable, knowledge-base adaptation, and voice-interaction trigger words. It also covers SkillHub standard installation prompts, typical conversational usage, and data timeliness notes.
Read MoreAll-in-One Financial Crawler: An Integrated Skill for Institutional Data, Public Sentiment, and Multi-Source Financial Queries
Introducing the SkillHub community skill "All-in-One Financial Crawler (with new MCP integration)," @user_9d5a2a39/cnfinancialscraper (v10.0.0). It supports a directory of 2,334 institutions, 60+ public opinion sources, A-share announcements and research reports, multi-source querying across 9 adapters, and scheduling via CrawlerDispatcher, with web enhancements through Shadow DOM/iframe/OCR and 79 MCP tools. It also covers `skillhub install` usage, the skills directories in various clients, and notes on `requires_api_key`.
Read Moretool-design: Designing Routable and Recoverable Tool Contracts for Agents
The **tool-design** skill from **Agent Skills for Context Engineering** focuses on the agent tool interface layer: it treats tools as contracts between deterministic systems and nondeterministic models, and guides tool descriptions and schemas, response formats, verb-noun naming and MCP fully qualified names, actionable error recovery, and pruning of overlapping tool sets based on consolidation principles. It is suitable for writing or debugging tool routing, consolidating tool catalogs, and designing MCP service boundaries; decisions such as project pipelines and sub-agent decomposition should be routed to other skills. The article is based on the official `SKILL.md` and covers enabling conditions, core principles, an audit checklist, and positive and negative examples, while explaining how it is loaded from a GitHub directory in `SKILL.md` format.
Read Moregcloud Skill: Constraining GCP CLI Operations for AI Agents
The gcloud Skill (v1.0.0) released by Google in the google/skills repository requires agents to validate syntax with leaf-level `gcloud help` before planning or executing `gcloud` commands, prohibiting the use of web searches to guess flags. It enforces minimizing list data, using single commands without pipes, `--quiet` and explicit `--project`/location, along with a four-step plan and dry-run. It covers both direct CLI usage and the MCP `run_gcloud_command`, explicitly listing operations such as IAM, deletion, and billing as prohibited from autonomous execution. Install it with `npx skills add google/skills`.
Read Morevoltagent-core-reference: VoltAgent Core Class Configuration and Lifecycle Reference
voltagent-core-reference is a reference Agent Skill maintained by the VoltAgent team for the VoltAgent class in @voltagent/core. It documents the constructor options, lifecycle order, and public methods, and marks deprecated fields. It is suitable for coding assistants to reference on demand when configuring MCP/A2A, separating memory between agents and workflows, connecting to a server or serverless runtime, or using VoltOps observability. Install it with npx skills add https://github.com/VoltAgent/skills --skill voltagent-core-reference.
Read Morenotion-meeting-intelligence: Automatically Prepare Meeting Agendas and Materials Using Notion Context
Maintained by OpenAI, notion-meeting-intelligence uses the Notion MCP to search for and retrieve pages, selects reference templates based on meeting type, creates or updates agendas and pre-reads in Notion, and can supplement research with Codex research. The article covers the Skill’s positioning, core capabilities, and the official Quick start, and explains how to install it with npx skills, configure the Notion MCP in Codex, and complete OAuth login. It is well suited for teams that have already built up documentation in Notion and want to reduce manual work in preparing pre-meeting materials.
Read Moredsh-mcp-scanner: An MCP Server Management Tool for DeepSeek Harness
This is a plugin for DeepSeek Harness (DSH) that simplifies scanning and managing MCP servers and their tools through a graphical interface. It supports full lifecycle operations (CRUD), multiple transport protocols, and host mount detection. The plugin depends on `@deepseek-ai/dsh-mcp-client`, stores configuration in `$DSH_HOME/mcp-servers.json`, and includes security mechanisms to protect local APIs. Developers can intuitively debug and configure MCP tools on the settings page.
Read MoreDeepSeek Harness Plugin dsh-nx: Secure Local Siemens NX Automation via MCP
The `dsh-nx` plugin for DeepSeek Harness (DSH) provides models with secure local Siemens NX automation capabilities via the MCP protocol. It supports part creation, sketching, feature generation, and STEP file export, and includes auxiliary tools such as health checks and state management. Installation is simple and prebuilt assets are supported. Note that this version is experimental: mock results are not written to actual files, and it must not be used in production environments. Developers can refer to the documentation for integration testing.
Read MoreDeepSeek Harness Plugin: BaizhiCloud Agent Toolkit Configuration Set
This is a community configuration package named `dsh-plugin-baizhi-agent-toolkit` for integrating the BaizhiCloud Agent Toolkit into DeepSeek Harness (DSH). The plugin reuses DSH's built-in MCP client, connects via Streamable HTTP endpoints, uses API keys for static authentication, and automatically discovers tools. Installation requires DSH 0.1.5-rc.2, and the configuration package does not include backend logic. When using it, environment variables must be configured, and tool names follow the `mcp__baizhi__<tool-name>` format. This plugin is an unofficial open-source project and is for community reference only.
Read Moreopencode-eyes: Giving Text Models a Pair of Eyes (StepFun step-3.7-flash)
The DeepSeek Harness plugin, opencode-eyes, provides visual capabilities for pure text large language models by mounting StepFun’s step-3.7-flash model to enable multimodal completion. The plugin packages an MCP server for a ready-to-use setup. To use it, users need to configure the `STEP_API_KEY`, with support for local image paths and custom prompts. The plugin allows adjusting parameters such as timeout, image size, and quality, while users should be mindful of runtime directory and permission limitations.
Read Moredsh-gitnexus-plugin: DeepSeek Harness Code Intelligence Knowledge Graph Integration
The `dsh-gitnexus-plugin` is a plugin for DeepSeek Harness (DSH) designed to help agents understand codebase structure. It uses GitNexus to index a codebase as a knowledge graph and provides `mcp__gitnexus__*` tools supporting concept search, context retrieval, impact analysis, taint discovery, change detection, and renaming. Installation requires Node.js >= 20 and the GitNexus CLI. The plugin spawns an MCP server over stdio transport and supports configurable timeouts under the MIT license.
Read MoreWin-Pilot: An Agent-Neutral Windows Computer Use Plugin
Win-Pilot is an agent-agnostic plugin designed to provide Windows agents with a unified `computer` tool, resolving the fragmentation between desktop application and browser control. It supports multiple integration methods, including MCP and CLI, and uses native control, UI Automation, and CDP to enable background operations and a recoverable protocol. Compatible with Windows 10/11 and requiring Node.js 22.12 or later, it occupies a dedicated DSH slot. Once installed, it enables unified management of system applications, windows, and browsers, with privacy protection and sandbox compatibility, significantly lowering the barrier to agent migration.
Read MoreAgent Web Search
DeepSeek Harness (DSH) introduces Agent Web Search, an open-source, provider-neutral search plugin that replaces the built-in tool. It provides agent-native and model-native grounding APIs with support for MCP, CLI, and Python. Installation adds the plugin via DSH, replacing the standard `web_search` capability.
Read Moredsh-mcp: MCP Management Interface and Tool Search
dsh-mcp is a DeepSeek Harness client plugin designed to resolve tool-list bloat and cache invalidation caused by integrating multiple MCP servers. It provides visual management, fine-grained tool-level control, process-level environment variables, and OAuth authentication support. To save tokens, it offers on-demand retrieval (search) and full injection modes, and supports image pass-through and system certificate validation. Users can install it from the command line and manage servers under Settings → MCP, with support for declarative takeover, OAuth flows, and Tool Search mode.
Read Morexby-nba-stats: NBA Data Service Plugin
xby-nba-stats is an MCP-based NBA data source plugin designed to provide real-time data interfaces for sports analytics agents. It offers comprehensive functionality, including basic score lookups, player and team information, search parsing, advanced statistics (such as PER and shooting heat maps), and league rankings. Users must obtain an API key from the official website and configure it, then activate it using the `set_xby_apikey` command. Once installed, the plugin supports querying schedules, rosters, and game logs, as well as performing complex analyses, making it suitable for NBA agent development.
Read Moreliudejua27-blip/fitmeet-dsh-plugin: Agent Social Networking
liudejua27-blip/fitmeet-dsh-plugin is a DeepSeek Harness plugin based on MCP 2.3, designed to address the pain point of agents connecting with strangers. It provides 23 tools that help an Agent discover resources, make plans, establish exchanges, and maintain connections, acting as a social connector. Users can issue natural-language commands to search the network, such as looking for a ball partner or a carpooling companion. Installation requires Node.js 22.19+ and OAuth 2.1 + PKCE authentication. The plugin transforms the Agent into a social agent, suitable for users who want to expand their social circles.
Read Moredsh-mcp-bridge: Exposing DeepSeek Harness as an MCP Server
`dsh-mcp-bridge` is a plugin that converts the DeepSeek Harness (DSH) toolset into a standard MCP server over HTTP. It enables external MCP clients, such as Claude Code, to securely invoke DSH capabilities through HTTP endpoints. The plugin runs on top of DSH's sandbox, supports streaming, requires zero dependencies, exposes only read-only tools (read, glob, grep, web_search) by default, and uses an allowlist mechanism to control risk. With simple configuration, it turns DSH into a tool provider within the MCP ecosystem.
Read Moredsh-mcp-workspace-scope: Directory-Based Restriction of MCP Tool Injection in DeepSeek Harness
`dsh-mcp-workspace-scope` is a plugin for DeepSeek Harness (DSH) designed to address the issue of MCP tools polluting the context window. It restricts tool injection using directory-level allowlists, applies the longest-path matching rule, and supports automatic inheritance by subdirectories. The plugin provides a dual rejection mechanism that not only hides tools but also intercepts their invocation. It also supports session-level temporary toggles and visual editing, allowing services to be enabled on demand in the Composer. After installation, restarting the Profile is required, and the default configuration file is located at `~/.dsh/mcp-workspace-scope.json`.
Read Morexby-ddg-search: DuckDuckGo Search and Content Scraping
xby-ddg-search is an MCP-based plugin designed to provide DSH with basic internet access. It supports DuckDuckGo search, web content fetching and parsing, and allows setting an API key. Users can install it via command line and must obtain a key from the Xiaobenyuan official website before use. The plugin runs with the current process permissions, suits agent scenarios requiring online queries, and its source code is licensed under MIT.
Read Morexby-cpan-package-info: A Tool for Fetching CPAN Package Metadata and README
`xby-cpan-package-info` is an MCP server plugin for DeepSeek Harness (DSH), designed to simplify interactions between agents and the Perl ecosystem. This plugin wraps CPAN query capabilities, supporting package information retrieval, dependency lists, README documentation, and search functionality. Before use, you need to register on the [Xiaobenyang official website](https://xiaobenyang.com) to obtain an API key and configure it using the `set_xby_apikey` command. The installation command is `dsh plugin --profile web add github:xby_skill/xby-cpan-package-info`; it is recommended to review the source code and license before use.
Read Morexby-constellation: Constellation Info Lookup and Horoscope Analysis Plugin
xby-constellation is a Model Context Protocol (MCP)-based astrology data query plugin developed specifically for DeepSeek Harness (DSH). It addresses the pain point of agents needing to maintain backend services to access domain-specific knowledge, providing complete capabilities such as basic zodiac information, rising sign calculation, daily horoscopes, and compatibility analysis. Users need to obtain an API key from the official website and configure it via `set_xby_apikey` to call various interfaces for data queries. Licensed under the MIT License, the plugin provides a standardized astrology analysis solution for the DSH ecosystem.
Read Morexby-celo-composer-kit: Celo Composer Kit Components and CLI Documentation Query
xby-celo-composer-kit is a DeepSeek Harness client plugin for macOS, designed to streamline the integration and development workflow of Celo Composer Kit. The plugin provides nine core capabilities via the MCP protocol, including API key management, component lookup (list, details, examples, search, and Props), category-based filtering, and CLI documentation queries. Developers can quickly obtain component information and usage guides through a conversational interface without repeatedly consulting documentation, thereby improving local development efficiency.
Read MoreMCP Server Management Panel for DeepSeek Harness (DSH) Web GUI
The DSH Web GUI plugin `dsh-mcp-skill-control` provides a graphical interface for managing MCP servers and local skills, eliminating the need for manual configuration file edits. It supports checking server status, starting, stopping, and restarting servers, adding servers (including via JSON import), deleting servers, and running diagnostic probes. Skill management includes one-click enabling or disabling of local skills and locating their files. All changes are written to `cordis.patch.yml` through YAML AST editing, ensuring configuration integrity. Installation requires the CLI, and uninstallation is equally straightforward. Developers should be mindful of version compatibility and build dependency issues.
Read Moredsh-docx-mcp: A Dependency-Free MCP stdio Service for Generating .docx Files
`dsh-docx-mcp` is a local, zero-dependency plugin for DeepSeek Harness that generates real `.docx` files. It supports headings, paragraphs, lists, and tables, and runs without installation. Agents can invoke it via the `docx_generate` tool by providing title, content, and structure parameters to directly output document files. Currently, the plugin requires manual path configuration, making it suitable for quickly generating contracts, reports, and similar documents.
Read Moredsh-mcp-oauth-client: General-purpose MCP Management and Client Bridge Plugin
`dsh-mcp-oauth-client` is a universal MCP management plugin for DeepSeek Harness (DSH). It addresses the complexity and error-proneness of native YAML configuration by providing a visual management interface. The plugin supports both HTTP (including streaming authentication methods such as OAuth 2.1 and Bearer tokens) and stdio (local process) service types, and offers capabilities for creating, reading, updating, and deleting services, as well as automated OAuth authorization and token refresh. It is installed via npm, and its configuration is written to a profile file. Users should be aware of security and compatibility considerations, such as plaintext credential storage and token persistence in memory.
Read Moredsh-skillhub: Unified Management of Local Skills and MCP in DeepSeek Harness
`dsh-skillhub` is a plugin for DeepSeek Harness designed to centrally manage local skills and MCP services. It supports independent toggles at the global, project, and conversation levels, with an intelligent downward propagation mechanism to ensure flexible and secure configuration. The plugin only controls visibility and does not modify files on disk. The installation command is `dsh plugin --profile web add @aa2246740/dsh-skillhub@1.0.7`, intended for the Web CLI and compatible with DSH 0.2.0-rc.2 or later.
Read MoreThe MCP Service Butler for DeepSeek Harness
The `dsh-mcp-market` plugin provides a visual MCP server management interface for DeepSeek Harness Web, aiming to replace cumbersome manual configuration. The plugin supports adding services from market templates (such as Apifox and Pixso) and also allows users to paste Claude Code-formatted JSON for manual configuration directly. New or modified services are automatically written to the local configuration file, with support for hot loading to take effect immediately, without requiring a restart. Users can enable, disable, and edit services in the panel. Additionally, the plugin registers MCP client instances by maintaining the `cordis.patch.yml` file, enabling the model to invoke the relevant tools directly. When uninstalling, attention should be paid to cleaning up the managed area.
Read Moredsh-tool-search: Progressive Streamlining of the DSH Tool Catalog
The DSH Harness plugin `dsh-tool-search` is designed to reduce token consumption caused by overly large tool catalogs. It adopts a "progressive disclosure" strategy, using the `tool_search`, `tool_describe`, and `tool_call` bridging tools to keep long-tail tools collapsed in the background. The plugin dynamically adjusts its hierarchical display based on the context budget, keeping core tools always visible and dynamically injecting context after tools are searched and invoked. It supports tool grouping through conversational configuration, with both static and runtime configuration options. It relies on a reranking matcher for precise search and uses an LRU limit to manage dynamically injected tools. Installation requires the DSH CLI and is well suited to projects using a large number of MCP servers.
Read Moredsh-vercel-mcp: Official Vercel MCP Connection for DeepSeek Harness
`dsh-vercel-mcp` is a plugin that provides an official Vercel MCP connection for DeepSeek Harness (DSH). It resolves the issue of the built-in client lacking support for OAuth 2.0 authentication by implementing dynamic client registration, PKCE verification, and automatic token refresh. Authorized through a web settings panel or an Agent command, the plugin exposes API features—such as deployments, project management, domains, and environment variables—as `mcp__vercel__*` tools within the agent session, enabling direct interaction with the Vercel platform.
Read MoreDeepSeek Harness Plugin: Connecting to Futu's Official MCP Server
`@yangzhe1991/dsh-futu-mcp` is a DeepSeek Harness (DSH) plugin that integrates Futu OpenAPI into agents, supporting financial data queries and trading. The plugin features secure authentication, permission control, deferred authorization, and zero-traffic startup. By default, it only grants read-only permissions; trading requires configuration adjustments. Tokens are securely stored locally and support automatic refresh. After installation, restart `dsh web` to trigger tool calls directly through conversation.
Read Morexby-wikimedia-search-images: Wikimedia Commons Image Search MCP Service
DeepSeek Harness (DSH) has released the online tool plugin `xby-wikimedia-search-images`, designed to address the issue that AI assistants cannot directly access Wikimedia Commons. The plugin runs as an MCP server and supports image search, metadata retrieval, and generating thumbnail collages for easier visual comparison. Its core features include API key setup and performing image searches. Developers can install it via the command line and use it after setting the key in chat. The plugin is suitable for scenarios requiring visual analysis or resource lookup based on Wikimedia assets, enabling developers to integrate high-quality image resources at low cost.
Read Morexby-time: Time Protocol Service for DeepSeek Harness (DSH)
xby-time is an open-source plugin for DeepSeek Harness (DSH), designed to provide standardized time services to agents via the MCP protocol. It includes three core functions: setting an API key, retrieving the current time in a specified time zone, and converting times across time zones. Users need to obtain an API key from the official Xiaoben Yang website and install it using the official command. The plugin is released under the MIT license, supports local debugging, and helps agents handle real-time time data conveniently.
Read Morexby-review-code: DeepSeek Harness Code Review Plugin
`xby-review-code` is a Model Context Protocol (MCP)-based code review tool server designed to provide DeepSeek Harness (DSH) with structured code review and scoring. It does not directly invoke large models; instead, it is responsible for constructing review prompts and parsing scores. The plugin offers core capabilities such as API key setup, overall code review, Git diff review, single-file review, and score extraction. Users must first register on the official website to obtain an API key, then install and configure it via setup commands, enabling integration into DSH workflows.
Read Morexby-osrs-stat: Real-Time OSRS Player Data and Leaderboard Service
xby-osrs-stat is an MCP server plugin for DeepSeek Harness (DSH) that provides real-time data services for *Old School RuneScape*. It allows agents to directly obtain comprehensive player data, skill leaderboards, and activity leaderboards, as well as perform player data comparisons. Users must first obtain an API key from the official website and configure it using `set_xby_apikey`, after which they can invoke features such as `get_player_stats` and `get_skill_leaderboard` in conversation. The plugin is open source and released under the MIT License, making it suitable for agent applications that need to integrate OSRS game data.
Read Morexby-ons-data: An MCP Server for Accessing UK Office for National Statistics Data
`xby-ons-data` is an MCP server plugin in the DeepSeek Harness (DSH) ecosystem for accessing the UK Office for National Statistics (ONS) Beta API. It simplifies the integration of official statistical data through standard commands—such as setting an API key, searching or retrieving datasets, and fetching observations—without requiring users to manage API keys manually. Users must first register on the official website to obtain a key, then install and enable the plugin via the DSH command line.
Read Morexby-skill/xby--rpc: OpenRPC-Based JSON-RPC MCP Server Plugin
`xby-skill/xby--rpc` is an MCP server plugin based on the OpenRPC specification, designed to enable the DSH large model to directly invoke remote JSON-RPC services. It provides three core functions: setting an API key, executing arbitrary method calls, and automatically discovering service capabilities. Users can interact by specifying the server address and method through conversational prompts. The plugin supports installation from GitHub, a local directory, or a cordis.patch.yml file; an API key must be obtained from the official website before use.
Read Morexby-glama-registry: DeepSeek Harness (DSH) Registry Service Search Plugin
The DeepSeek Harness (DSH) plugin `xby-glama-registry`, maintained by xby-skill, is designed to address the lack of tools in agent development. It allows developers to search for servers through the Glama MCP registry and provides features for setting an API key and searching for matching servers (such as 'telegram' and 'docker'). Users must first obtain an API key from the official website and install the plugin using the `dsh plugin` command. The project is released under the MIT license and is intended for local or personal use only.
Read Morexby-gathering-card-lookup-service: Magic: The Gathering Card Lookup Service
The DeepSeek Harness plugin `xby-gathering-card-lookup-service` is based on the MCP protocol and provides DSH agents with Chinese Magic: The Gathering card lookup capabilities. It supports searching card details by set code and number, as well as set data queries and a randomized shuffle feature. Users must obtain and configure an API key from the official website before using various syntaxes (such as `t:creature`) for flexible card search and filtering.
Read Morexby-fastdomaincheck: A Bulk Domain Registration Status Checker Based on the MCP Protocol
`xby-fastdomaincheck` is a domain registration status check server based on Python and the MCP protocol, maintained by `xby-skill`. This plugin supports automatic API key management and bulk domain checking. Users must obtain the API key from the official website and install the plugin using the `dsh plugin` command. It is suitable for scenarios that require domain availability validation in agent development.
Read MoreDeepSeek Harness Plugin: xby-dream-of-red-chamber-relationship
This article introduces a *Dream of the Red Chamber* character relationship graph plugin designed specifically for DeepSeek Harness. Through the MCP tool interface, the plugin converts graph data structures into callable tools, with core features including setting API keys, retrieving the database schema, and executing read-only Cypher queries. Users must first obtain an API key from the official website, after which they can directly interact with the graph database in a conversational environment, facilitating the construction of agents based on structured knowledge. The plugin is released under the MIT license and supports command-line installation and activation.
Read Morexby-daily-hot: An MCP-Based All-in-One Aggregation Service for Web-Wide Trending Topics
xby-daily-hot is a DSH plugin based on the Model Context Protocol (MCP), designed to simplify the process for agents to access trending data across the entire web. The plugin provides a one-stop trending topic aggregation service, covering news and information, social media, technology, and video communities. It supports mainstream platforms such as Baidu, Weibo, and Bilibili, and offers custom RSS and web scraping capabilities. After registering on the official Xiaobenyang website to obtain an API key, users can quickly invoke various types of trending data through a unified command, making it suitable for scenarios such as news aggregation analysis and agent development.
Read Morexby-berlin-search-services: Berlin Administrative Services Search Plugin
The DeepSeek Harness plugin `xby-berlin-search-services` is designed for developers to enable agents to access Berlin administrative services data. It provides features such as API key management, service search, detail lookup, list retrieval, and data statistics. Users must first install it and obtain an API key using the official command, then configure the key in the DSH interface to retrieve information on over 1,000 public services offered by Berlin authorities using tools such as `search_services` or `get_service_details`. This tool is maintained by `xby-skill` and is released under the MIT license.
Read Morevalkia/dsh-plugin-computer-use: Computer Use Plugin for DeepSeek Harness
The DeepSeek Harness (DSH) plugin `valkia/dsh-plugin-computer-use` enables DeepSeek models to control the computer interface through the Open Computer Use MCP protocol. It is designed to address the limitation of models being unable to directly operate a computer and is categorized under “model reasoning.” Developers can use this plugin to let DeepSeek agents perform computer operations in local or remote environments. The plugin runs with the permissions of the current DSH process, so the source code and license should be reviewed before use. See the related repository for more details.
Read Moredsh-mcp-plus: A GUI Management Plugin for the DSH MCP Server
DSH MCP Plus is a graphical management plugin developed by sojo-negai, designed to simplify the configuration and monitoring of MCP servers in DSH. It supports editing JSON configurations directly in the web interface, viewing real-time connection status via green, yellow, or white dots, independently starting or stopping registered tools, and automatically reconnecting them. It is also compatible with the Claude Desktop format. Installation requires running the pnpm command in the project root directory, after which you can refresh the page to start using it.
Read Moredsh-wall-mcp-manager: Centralized Management of MCP Service Configuration and Tool Contracts
`sduwall/dsh-wall-mcp-manager` is a plugin in the DSH ecosystem designed to address the complexity of managing multiple MCP services. It automatically reconciles configuration manifests to mount, update, and unmount services, applying changes without restarts. The plugin provides centralized configuration management, visualized tool contract display, and credential masking (showing only the configured status) as security features. Its interface supports local loopback access only, and both credential values and key names are never exposed over external networks. Through strict permission and security boundary design, it ensures both operational convenience and security when dynamically adjusting service states.
Read MoreDeepSeek Harness Plugin: mikotomywife/dsh-mcp-loader - Lazy Loading and Session-Based Masking for MCP Tools
`dsh-mcp-loader` is a lazy-loading plugin for MCP tools in DeepSeek Harness (DSH), designed to address the token consumption and tool-selection difficulties caused by a large number of MCP servers. The plugin introduces a "loader" mechanism whereby actual tools enter the model context only when invoked, eliminating fixed overhead and streamlining the tool list. Key features include loader switches isolated per server/session (supporting `auto`, `lazy`, and `eager` modes), session-scoped masking to hide tools, global tool disabling, description engineering optimization, environment cleanup, and resource bridging. It supports atomic registration, fault tolerance, and permission/security checks, significantly improving model selection efficiency in multi-tool scenarios.
Read Moredsh-mcp-skills: Skills and MCP Management Plugin for DeepSeek Harness
dsh-mcp-skills is a DeepSeek Harness plugin designed to centrally manage skill files and MCP server configurations through a Web interface. It supports browsing, importing, and deleting skills, as well as configuring, testing, and mounting MCP servers. Its core advantage is transforming scattered YAML configurations into visualized CRUD operations, supporting both stdio and HTTP types while ensuring the atomicity and safety of file operations. Installation requires the use of the pnpm environment variable, and the service must be restarted afterward.
Read MoreSiYuan Codex Bridge: Local SiYuan MCP Endpoint and DeepSeek Harness Plugin
The DeepSeek Harness plugin aims to integrate SiYuan Note into DSH via the MCP protocol, enabling “everything as a plugin.” Users can install it through the desktop app or command line, and configure startup control, operation permissions (read-only / creation / full access), and the token in `config.json`. The plugin registers SiYuan’s 29 official capabilities as `mcp__siyuan__*` tools, supporting natural-language operations and AI tool groups, including conversation, editor actions, and agent features. It does not write to the plugin directory, stores audit logs at the specified path, requires Node >= 22, and provides diagnostic commands.
Read Moredsh-algovault: AlgoVault MCP Server Mount Configured for DeepSeek Harness
dsh-algovault is a plugin for DeepSeek Harness designed to streamline the integration of external MCP services. It provides capabilities such as one-click mounting, composite trading signals, market condition detection, and cross-exchange arbitrage. Installation requires adding it via a command and restarting the configuration. After enabling an API key, the model can use its provided tools to generate strategic recommendations. When using it, note that the model should treat HOLD signals as stop instructions and that skill files must be loaded manually.
Read Moredsh-mcp-tunnel: Spinning Up an Outbound Tunnel for a Local MCP Server
`dsh-mcp-tunnel` is a networking tool plugin for DeepSeek Harness (DSH) designed to address the issue of local MCP Servers being inaccessible to remote Agents. It creates secure tunnels via a Docker Compose stack, using `mcp-proxy` and `cloudflared` to map local services to HTTPS URLs without exposing local ports. The plugin supports automatic configuration generation, certificate management, and CLI commands, but does not automatically modify host configurations; users must manually add the generated configuration snippets. Before use, a Docker environment is required, and production deployments must include their own authentication layer.
Read MoreHuayu Yuandian Legal Data (huayu-yuandian-legal-data) - DeepSeek Harness Official Plugin
`huayu-yuandian-legal-data` is an official DeepSeek Harness plugin for accessing Chinese-language legal data. The plugin provides four categories of official Skills through an aggregated MCP entry point: regulations, cases, enterprises, and securities compliance. It supports OAuth and API Key authentication and includes connection stability safeguards. Users can install and configure it via CLI or the settings page, and invoke registered MCP tools directly in DSH. Currently, the plugin supports tool bridging only and provides a data disclaimer and an MIT license.
Read Moredsh-wind-aifin: Wind AIFin Integration in DeepSeek Harness
`xiamu-ssr/dsh-wind-aifin` is a security plugin for DeepSeek Harness that integrates the official Wind AIFin services into DSH, decoupling credentials from model Shell processes. The plugin connects to seven Wind Streamable HTTP MCP domains via DSH's `@deepseek-ai/dsh-mcp-client` and exposes them as native tools, while also providing the `wind_alice` tool for invoking financial analysis workflows. The agent does not directly handle API keys; credentials are managed by DSH and resolved per request. The plugin adapter listens only on `127.0.0.1`, accepts only fixed endpoints, and must not be used as a general-purpose proxy. The installed version must
Read Moredsh-matlab-mcp: Registering MATLAB/Simulink to DeepSeek Harness
`dsh-matlab-mcp` is a plugin for DeepSeek Harness (DSH) designed to inject MATLAB/Simulink control capabilities into the DSH GUI by registering a MathWorks MCP server. This plugin supports core functions such as code execution, static analysis, Simulink model viewing and editing, and Gherkin-syntax-based testing. Before use, you need to install the MATLAB Agentic Toolkit and pnpm, and meet the corresponding version requirements. After installation, restart the GUI and initialize a session in MATLAB; the agent can then interact with MATLAB through the `mcp__matlab__*` interfaces.
Read MoreDeepSeek Harness Plugin: dsh-mcp-diff
`fakek0f3st/dsh-mcp-diff` is a client-side plugin designed for the DeepSeek Harness (DSH) Web interface to provide a unified view of file changes. It supports MCP file system operations, built-in tools, and various Bash commands (such as sed, tee, and redirection), parsing changes into structured, collapsible diff cards that allow viewing added and removed lines and navigating to the IDE. The plugin generates visual diffs by analyzing the agent's file operation commands. For unrecognized commands or dynamic paths, it preserves terminal output and flags destructive operations. The plugin relies on platform modules and is compatible with DSH 0.1.1-rc.2.
Read MoreDeepSeek Harness Plugin: Knowz (Hosted MCP + Knowledge Base Skills)
Knowz is an MCP plugin that provides managed knowledge base connectivity for DeepSeek Harness (DSH). It focuses on knowledge retrieval and management, connecting agents to Knowz vaults via the MCP protocol and does not include code execution capabilities. Users need to install the plugin and configure an API key that starts with `ukz_`, then run the `knowz-setup` skill to map vaults. The plugin also offers multiple skills, including search, save, browse, and automatic management, while requiring attention to key security and registration pitfalls.
Read Moremcp-deliverable: DeepSeek Harness MCP Server Configuration Component
`mcp-deliverable` is a community Bundle for DeepSeek Harness, designed to address the cumbersome and error-prone process of configuring MCP servers through a visual Web interface and remote control capabilities. The plugin supports reading, atomically writing, and deleting configurations, and displays service registration status. It provides extensive editing options, including parameters, environment variables, working directory, and reconnection policies, and leverages the `dsh.bundle.patch` mechanism for automatic integration. Installation requires a specific Node.js version, and pinning a commit is recommended for production environments. Pay attention to Bundle priority and the secure storage of sensitive information.
Read MoreDeepSeek Harness Plugin: dsh-mcp-toggle
dsh-mcp-toggle is a DeepSeek Harness plugin developed by Zenjibad, designed to address the cumbersome management of MCP servers and the need for restarts. It adds an “MCP Servers” page to Settings, allows users to toggle server enabled/disabled states in real time, and persists the configuration. The interface adapts to light and dark themes, and critical system entries are safely protected. Users can install it via the specified command, with configuration changes taking effect immediately without restarting the DSH process.
Read Moredshx: MCP Client Hub and DeepSeek Harness Extension Management Suite
The DeepSeek Harness plugin, dshx, is designed to address the cumbersome configuration overhead in Web GUI development based on the “everything is a plugin” philosophy. It provides a unified Web management console that integrates an MCP client hub, memory store management, skill and slash command definitions, and prompt management. dshx supports multiple MCP protocols (stdio, Streamable HTTP, and SSE), and features automatic reconnection and real-time logging. Memory management supports both global and project-level scopes while adhering to a session sandboxing policy. Skills and commands are file-based and take effect immediately. Implemented in pure ESM JavaScript without a build step, the plugin securely stores credentials and binds only to the local address, making it well suited for developers who need to manage complex agentic capabilities in a unified manner.
Read Moredsh-netcatty: Connect the DeepSeek Harness to Netcatty's External MCP Server
`dsh-netcatty` is a plugin that connects DeepSeek Harness (DSH) to the Netcatty External MCP server. It allows DSH to securely invoke Netcatty's SSH and SFTP tools via standard protocols without exposing sensitive credentials. Core features include connecting to official servers, tool registration, permission control, and reconnection after disconnection. Installation requires DeepSeek Harness 0.1.0-rc.7 or later and Netcatty 1.1.80 or later, with support for non-standard path configuration and customized parameter adjustments. By default, it follows the principle of least privilege, and all authentication is handled natively by Netcatty. The plugin provides developers with a secure and convenient integration between DeepSeek and SSH/SFTP.
Read Moredsh-plugin-zai-mcp: Integrating the z.ai MCP Server into DeepSeek Harness
`dsh-plugin-zai-mcp` is a plugin-based extension designed to enable DeepSeek Harness (DSH) users to access z.ai's GLM Coding Plan capabilities without manual integration. It encapsulates z.ai's four MCP servers into 13 native DSH tools, covering visual analysis, web reading, search, and document reading. Installation supports Git sources, is zero-config by default, and can automatically reuse DSH's GLM Key. The plugin relies on z.ai cloud services; tool calls consume GLM Coding Plan quota and are subject to specific restrictions on media file sizes and code repository access permissions.
Read MoreDSH Plugin: MCP Management
DSH adopts an “everything is a plugin” architecture. The MCP management plugin maintained by Modole is designed to simplify the configuration of MCP Servers in DSH Web. This plugin supports managing both stdio and Streamable HTTP server types. Users can install the plugin by specifying a Git repository URL, and should be aware that it runs with current user privileges and that source code should be reviewed from a security perspective. After installation, restarting `dsh web` makes the plugin available, enabling direct management of server configurations through the Web interface.
Read MoreDeepSeek Harness MCP Server Plugin: meltartica/dsh-mcp-servers
`meltartica/dsh-mcp-servers` is a DeepSeek Harness plugin designed to simplify MCP server management. It supports stdio, HTTP, and SSE transport protocols, and provides automatic tool registration, real-time status monitoring, automatic reconnection, testing, and editing features. In addition, the plugin supports local JSON backup and recovery as well as WebDAV cloud synchronization, with an option for daily automatic backups. After installation, users can configure servers and verify connections in Harness settings. Note that tool schemas support only a whitelisted subset, and WebDAV is limited to HTTPS.
Read Moredsh-netcafe: Bringing the AI NetCafé Tool to DeepSeek Harness
`dsh-netcafe` is a DeepSeek Harness (DSH) plugin that extends functionality by injecting `@deepseek-ai/dsh-mcp-client` through configuration files. The toolkit includes invoice extraction, reconciliation statement checking, SQL dialect conversion, reachability monitoring for URLs in China, and resume review. All computations are completed deterministically on local servers without relying on external LLMs. It provides either the full catalog or four focused packs to control context costs, and supports executing operations such as batch invoice processing and creating monitoring tasks via simple JSON calls.
Read MoreDSH Plugin: dsh-skill-mcp-manager (Capability Manager)
dsh-skill-mcp-manager is a plugin for DeepSeek Harness Desktop (DSH), designed to address the cumbersome configuration management of MCP servers and Skills through a unified interface. It supports CRUD operations for skills, symbolic link synchronization, and folder synchronization, as well as session-based progressive loading and hot updates for MCP servers. By appending a capability list to the system prompt, it optimizes resource usage. The plugin requires DSH's profile bundle patch feature to work and takes effect after a restart following installation.
Read Moredsh-cordis-universal-adapter: A Bidirectional MCP Bridge Adapter for DeepSeek Harness
`dsh-cordis-universal-adapter` is a connector for the DeepSeek Harness (DSH) Host composition layer, designed to enable bidirectional interoperability between DSH and the external MCP ecosystem. This plugin primarily addresses resource loading and capability exposure on both sides. It supports loading external MCP servers (stdio/HTTP) and Agent plugins, registering them as native tools for Agent invocation; it also supports exporting DSH tools as an MCP server for use by other clients. The plugin includes features such as connection state management and semantic routing. Installation requires Node.js >= 22.19 and a specific version of DSH. Users should be aware of the security risks associated with running external tools and ensure API version compatibility.
Read MoreDeepSeek Harness: A Visual Management Plugin for the MCP Protocol
dsh-plugin-mcp-connector is a visual management plugin for DeepSeek Harness (DSH), maintained by developer gxinxing. It addresses the difficulty of managing MCP server configurations in the web interface by providing a graphical user interface via the `/mcp` path. It supports dynamically adding, deleting, enabling, or disabling MCP services, and features automatic tool registration, invocation log viewing, and configuration persistence. Users can install and enable it with simple commands, making it easy to manage external tool connections in a browser.
Read Moredsh-mcp-setting: Manage DSH MCP Servers in the Settings Interface
`dsh-mcp-setting` is a client plugin for DeepSeek Harness (DSH) that enables visual management of MCP server configurations through a Web interface. It supports CRUD operations, real-time monitoring of connection status, enable/disable control, and configuration validation. The plugin employs security mechanisms to ensure the atomicity and security of configuration changes, allowing them to take effect immediately via hot reload or restart. It is recommended to install it via npm into a specified profile, enabling dynamic adjustments of server parameters without restarting the process.
Read Moredsh-arcgis-pro-bridge: Connecting ArcGIS Pro to DSH via MCP
`dsh-arcgis-pro-bridge` is a plugin for DeepSeek Harness (DSH) that bridges ArcGIS Pro Python logic into DSH, enabling models to directly operate on local GIS data through standard MCP tools. It is pure ESM with vendored Python, has no build step, and supports installation via GitHub. Its core features include structural reading, spatial analysis (such as Buffer and Clip), and execution of custom ArcPy code. It requires a Windows environment, and ArcPy runs within the Python bundled with ArcGIS Pro without polluting the Node.js environment. Users should note that it is limited to local execution and that `execute_arcpy_`
Read More《Cursor Docs》-MCP Install Links
Looking to share MCP servers, rules, and more? Plugins make it easier to bundle and distribute everything in one package. You can publish to the Cursor Marketplace or your team's private marketplace.
Read Moreawesome-ios-sim: Turn iOS Simulator state into versionable code
# Introduction In DeepSeek Harness (DSH), integrating an iOS simulator with an agent typically involves opening a shell for the agent to directly call `xcrun simctl`. The practical challenges of this approach are significant: environment configurations are scattered across scripts and undocumented `defaults` commands, making test environments difficult to reproduce. Additionally, the model is exposed to an untyped and unsafe shell interface, making it difficult to audit what was changed and in what order. `qubyyang/awesome-ios-sim` solves this problem. It does not add new simulator capabilities but encapsulates "changing simulator state" into a fixed process: ```text profile + current snapshot ```
Read Moredsh-mcp-adapter: Enables DeepSeek Harness to invoke MCP on demand instead of injecting all schemas into the context
# Introduction Connecting MCP servers to DeepSeek Harness (hereinafter referred to as dsh) is most directly achieved via the official `@deepseek-ai/dsh-mcp-client`. This client connects to the server at startup and registers each declared tool as a native `mcp__<server>__<tool>` function. The trade-off is the schema cost: with multiple servers, hundreds of tool definitions are passed into the context with every request, regardless of whether they are utilized. `dsh-mcp-adapter` provides an alternative solution to this problem by registering only a single `mcp` proxy tool. It employs lazy server startup, allowing the model to `search` or `describe` tools on demand.
Read Moredsh-mcp-sync: Scan local MCP configurations and register tools for direct model invocation.
<tool_call>Bash<arg_key>command</arg_key><arg_value>ls -la "D:\D-Users\D-Administrator\.zcode\workspace\default" | grep -E "obj_|arxiv"</arg_value><arg_key>description</arg_key><arg_value>List obj* and arxiv related directories</arg_value></tool_call>
Read Moredsh-mcp-xcode: Integrate Xcode headless capabilities into DSH agent
# Introduction When developing macOS/iOS applications in DSH (DeepSeek Harness), there is a practical challenge: tasks such as building projects, running unit tests, viewing SwiftUI previews, and operating simulators are typically performed within the Xcode UI, preventing agents from intervening. DSH's philosophy is "everything is a plugin," and such capability gaps are usually filled by plugins. Starting with Xcode 27, the system comes with built-in headless MCP services (`xcrun mcp-server` / `xcrun mcpbridge`) that expose the aforementioned capabilities in the form of MCP tools. However, to use them within DSH, a bridging layer is required to convert these MCP tools into...
Read MoreDeepSeek Harness Tool Palette: Progressive Tool Discovery and Agent-based Unlocking
# Introduction The philosophy of DSH is "everything is a plugin." However, as more components are installed, a practical problem arises: every plugin and MCP server contributes a batch of tool schemas to each model request. As the list grows, the cost of redundant tool prefixes increases, and it becomes more difficult for the model to identify the correct target within a long list of tools. DeepSeek Harness Tool Palette adopts the opposite approach: when an Agent starts, it carries only a `tool_search` function and a set of base tools selected during deployment. Other globally installed tools remain hidden initially; they are unlocked only when the model searches for the corresponding capabilities, allowing them to be used in subsequent steps. The following sections explain the specific mechanism, installation process, and usage boundaries of this plugin. # What is this? Tool Palette by li...
Read Moredsh-jiey-browser: Enable DeepSeek Harness to use real Jiey Browser via MCP
**Introduction** When integrating browser capabilities into DeepSeek Harness (DSH) agents, the typical approach is Playwright: a headless environment with a separate automation profile. The login state and cookies are completely isolated from the user's daily browser, and installation requires running `npx playwright install` first. If a task requires reusing sites that are already logged in, this path involves a lot of extra work. dsh-jiey-browser takes a different approach: it hands over the real Jiey Chromium (i.e., the user's own browser) to the agent via MCP, rather than spawning a separate headless environment. The following sections introduce what this plugin can do, how to install it, and how to use it. # What is this? dsh-ji...
Read Moredsh-apple-mode: Connect Xcode AI stack to DSH sessions using agent presets
# Introduction If you are writing Swift on macOS using DeepSeek Harness (dsh), you have likely encountered this awkward situation: the model can generate SwiftUI code, but cannot access the Xcode project itself—creating targets, modifying build settings, switching schemes and run targets, reading diagnostic information, and editing the String Catalog—all of these tasks require you to manually switch back to Xcode. The philosophy behind dsh is that everything is a plugin, and since Xcode 26, Apple has officially provided `mcpbridge`, allowing external tools to directly manipulate open projects. This article introduces jihongboo/dsh-apple-mode, which brings this capability to DSH.
Read MoreQichacha MCP OAuth Plugin: Connecting to Qichacha MCP Service in DeepSeek Harness
`duhu2000/qcc-mcp-oauth` is an MIT licensed plugin for DeepSeek Harness (DSH) designed to simplify the integration of Qichacha MCP services. It automatically connects to Qichacha via one-click OAuth authorization (using Authorization Code + PKCE), supports six core data servers such as company and risk, and dynamically mounts resources based on token permissions. The plugin features automatic token refresh, local persistent storage (permissions 0700), and automatic recovery upon restart. Users can execute connection, view status, or disconnect authorization through conversation commands; the latter revokes the refresh_token and disables the tool. Installation requires DSH CLI and Node 20+, run `dsh plu...
Read Moredsh-plugin-manager: DeepSeek Harness Plugin, MCP, and Skill Management
dsh-plugin-manager is a plugin management system for DeepSeek Harness, maintained by CHF-hub99 and licensed under the MIT License. It aims to centrally manage plugins, MCP servers, and Skills within the DSH ecosystem, supporting status viewing, installation and uninstallation, enabling and disabling, and configuration import via the Web GUI and Agent tools. The Web interface allows users to install plugins by package name, add MCP services with various protocols, and upload and verify Skills packages; the Agent tools provide commands such as `plugin_install` and `mcp_add` for operations within conversations. The system persists configuration by modifying `cordis.patch.yml` and exposes an HTTP API accessible only via loopback. Installation command...
Read Moredsh-hud: DSH Web Floating Status Panel
dsh-hud is a community HUD plugin for the DeepSeek Harness Web interface: the input bar dashboard button opens a right-side floating panel, centrally displaying Git status and diffs, MCP services, Skills, model and token usage, session statistics, and account balance. Installable via `dsh plugin --profile web`, it is read-only and requires no build, making it suitable for quickly confirming repository and session status within the chat interface.
Read Moredsh-plugin-garmin-connect: Integrating Garmin Connect Sports and Health Data into the DSH Agent
dsh-plugin-garmin-connect (v0.1.5, maintained by Likenttt) is a Garmin Connect integration plugin for DeepSeek Harness. It registers activities, sleep, steps, heart rate, weight, training libraries, and other data as 10 AI-callable tools, supporting FIT downloads and training creation. Installed via `npx @deepseek-ai/dsh plugin add`, it allows natural language queries in web dialogs or standalone MCP clients after configuring the GARMIN environment variable. Note that version 0.1.5 does not yet support browser-based two-factor authentication; the plugin runs with dsh process permissions, so review the source code and MIT license before installation.
Read Moredsh-harness-mcp-server: Exposing Harness Capabilities as an MCP Service
dsh-harness-mcp-server is a DSH workflow plugin released by community maintainer chushixixin, which exposes the DeepSeek Harness agent runtime as a StreamableHTTP MCP service. External MCP clients (such as Hermes) can drive the Harness to execute real coding tasks like bash commands and file system operations through tools like agent_run and task_inbox, and retrieve structured results including changes and verification. The plugin listens on 127.0.0.1:8090 by default, supports npm installation and activation via cordis.yml patches, and is suitable for alternative scenarios requiring context isolation or parallel execution.
Read Moredsh-oomol: Progressive MCP Discovery for Invoking OOMOL Connector Actions in DeepSeek Harness
dsh-oomol is a DeepSeek Harness web connectivity tool plugin maintained by oomol-lab, which connects the Harness to the OOMOL Connector (either OOMOL Hosted or self-hosted OpenConnector). The Harness only stores the Connector client keys, while Provider credentials remain on the Connector side; Actions are exposed on-demand through progressive MCP discovery. This article introduces the installation commands, dual-mode configuration, typical conversational usage, and security boundaries, along with applicable scenarios and troubleshooting tips.
Read Moredsh-mcp-manager: Visually Managing MCP Servers in DSH
dsh-mcp-manager is a client plugin developed by Js2Hou for DeepSeek Harness, which lists installed MCP servers in the "Settings → MCP" section. It supports adding, removing, enabling/disabling, editing, and connection testing for MCP servers, while displaying real-time connection status and the number of tools. Configuration changes are written back to cordis.patch.yml via HMR, eliminating the need for manual YAML editing. The plugin can be installed using the command `dsh plugin --profile web add @js2hou/dsh-mcp-manager` or through community marketplaces such as dsh-market and DSH Desktop. Licensed under MIT, it is ideal for DSH users who need centralized management of multiple MCP servers.
Read Moredsh-chatgpt-bridge: Create and Supervise DSH Sessions with ChatGPT Web
dsh-chatgpt-bridge is an independent DSH plugin (v0.4.1, MIT) maintained by jiezeng2004-design. It enables ChatGPT Web to connect to the local bridge via MCP through a secure tunnel (default 127.0.0.1:3456/mcp), allowing the creation, viewing, continuation, and supervision of DSH agent sessions. It exposes 15 tools, including Goal supervision and approval. The installation command is `dsh plugin --profile web add dsh-chatgpt-bridge`, requiring Node.js >= 22 and sharing the web profile with `dsh web`. The control flow is from ChatGPT to DSH, without altering the DSH core security model.
Read Moredsh-mcp-manager: A MCP Management Plugin for DSH with OAuth and stdio Capabilities
dsh-mcp-manager is a DSH web plugin maintained by hyqhyq3, managing MCP servers on the Settings → MCP page. It supports HTTP OAuth (PKCE + dynamic client registration), static tokens (environment variable names not stored on disk), and stdio local processes, with tools registered under the mcp__<name>__* namespace. Optional features include on-demand proxy schema compression, support for workspace-level mcp.json, and exclude blocking. Installation: npx -p @deepseek-ai/dsh dsh plugin --profile web add github:hyqhyq3/dsh-mcp-manager.
Read Moredsh-search-mcp: Using Search MCP to Replace DSH's Built-in Web Search
dsh-search-mcp is a DSH networking tool plugin maintained by gxpppp. While retaining the web_search tool name and display on the model side, it redirects search requests to Tavily, Brave, Exa, Perplexity, DuckDuckGo, or custom MCP servers, and disables the built-in DeepSeek search provider. It is compatible with DSH 0.1.0-rc.7 and Node.js 20+, allowing configuration of servers, credentials, and result counts through a web settings page. This article covers installation links, credential configuration, provider presets, combination verification, and uninstallation and restoration methods.
Read Moredsh-reef: One-Time Installation for Browser, MCP, and Code Hosting Integration
dsh-reef is a DSH plugin bundle maintained by huey1in, offering five modules with a single installation: native embedded panels, 21 browser_* browser tools, MCP Server, GitHub/GitLab read-only tools, and webhook automatic reviews. It supports multi-tab, multi-profile browser sessions, exposing DSH sessions to external MCP clients, automatic PR/MR reviews, and issue auto-fixing loops. The installation command is dsh plugin --profile web add dsh-reef, licensed under MIT, and requires Node.js >= 22.
Read MoreLocal Shell MCP: Enabling Controlled Shell and File Workspaces for DSH
local-shell-mcp is an MCP control plane maintained by fwerkor, offering clients controlled shell, file, browser automation, and remote worker capabilities. When used as a DeepSeek Harness plugin, it fully bridges the mcp__lsm__ tool interface via HTTP, binding each DSH session to an independent logical session and live workspace. The article introduces the recommended topology, steps to start LSM before installing the plugin, environment variable configuration, and typical usage, and explains that the plugin runs with DSH process permissions, requiring verification of source code and security boundaries before installation.
Read Moredsh-skill-mcp-panel: Unified Management of Skill and MCP in DSH Web Settings
dsh-skill-mcp-panel is a DSH client plugin maintained by Fishquito7, offering two management panels—Skill and MCP—within the web settings page, accompanied by the dsh-panel command line. Skills support workspace partitioning, grouping, batch migration, and scoped enablement/disablement, while MCP enables configuration writes to the cordis.patch.yml managed block via Stdio/HTTP with hot-reloading. It is recommended to install via the v2.0.1 release tarball; after restarting the gateway, the features become accessible on the settings page.
Read MoreArgo: Multilingual Search and Evidence Verification for Agents
Argo (taxuseek/argo) is a multilingual search infrastructure specifically designed for AI Agents, routing queries to 120+ engines by domain and language, and producing concise JSON with confidence scores instead of link lists. It supports one-click installation via DSH, MCP mounting, and command-line invocation, covering vertical scenarios such as finance, film, academia, and code. It is ideal for intelligent agent developers requiring auditable and token-efficient web-connected capabilities.
Read More《Cursor Docs》-MCP Install Links
Looking to share MCP servers, rules, and more? Plugins make it easier to bundle and distribute everything in one package. You can publish to the Cursor Marketplace or your team's private marketplace.
Read MoreThermal Management of MCP and Skills in the Settings Panel via dsh-plugin-manager
Li Qichen's maintained dsh-plugin-manager integrates MCP switch/delete, Skills recycle bin and built-in plugin package browsing into the DeepSeek Harness settings panel, automatically backs up before modifying cordis.patch.yml, and takes effect via host HMR hot update without restarting the dsh web. This article checks the installation commands, capability boundaries and precautions against the community directory page, repository README and source code. The community directory is not an official store, and plugins run with the current dsh process permissions.
Read MoreInstall verified MCP servers for DeepSeek Harness at once with dsh-mcp-bridge
**Translation:** dsh-mcp-bridge is a DeepSeek Harness community plugin maintained by Edge-Echo, which packages the official @deepseek-ai/dsh-mcp-client and a curated set of MCP servers into an installable bundle. By default, only the "everything" demo is enabled, with presets annotated for memory, filesystem, GitHub, Playwright and remote HTTP services. A verify script and CI connectivity checks are added via the servers/ directory definition. The project is categorized under the memory taxonomy. This article covers the directory page, cross-verified installation commands from GitHub and npm, capability boundaries and precautions.
Read MoreManage DeepSeek Harness's MCP servers with dsh-mcp-panel
### 翻译版本1（贴合技术文档正式译法）： DSH-MCP-Panel is a DeepSeek Harness interface enhancement plugin maintained by PerryLink, built upon the official `@deepseek-ai/dsh-mcp-client`: It allows users to check server statuses, tools, errors and reconnection counts via the `/mcp` command and the MCP tab in the settings page, with output desensitized. All add, delete and modify operations are completed via append-only `cordis.patch.yml` snippets, and approvals are written with automatic backups. This article verifies the installation commands, capability boundaries of version 0.4.0 and relevant precautions based on the community directory, GitHub and npm. ### 翻译版本2（更口语化适配社区语境）： DSH-MCP-Panel is a DeepSeek Harness UI enhancement plugin maintained by PerryLink, built on top of the official `@deepseek-ai/dsh-mcp-client`. You can use the `/mcp` command or the MCP tab in the settings page to view each server's status, tools, error logs and reconnection times, with all outputs desensitized. All create, edit and delete operations are done through append-only `cordis.patch.yml` fragments, and approvals will be saved automatically after being written. This article confirms the installation commands, the capability limits of version 0.4.0 and the matters needing attention based on community directories, GitHub and npm.
Read MoreIntegrate OAuth 2.1 MCP into DeepSeek Harness with dsh-oauth-mcp-client
`dsh-oauth-mcp-client` is a DeepSeek Harness community plugin maintained by Springbrand-lab. It adds OAuth 2.1 support (authorization code + PKCE, dynamic client registration, browser login and local callback) based on the native `dsh-mcp-client`. It connects to the Springbrand production MCP Gateway by default, and users can also add other services that support OAuth and Streamable HTTP via the web interface. Tokens are stored by the DSH credential service. It is licensed under MIT, and currently needs to be built locally before being installed into the web profile.
Read MoreManage MCP Servers in DeepSeek Harness Settings Page with dsh-mcp-manager
`dsh-mcp-manager` is a DeepSeek Harness community plugin maintained by hyqhyq3, which manages remote HTTP and local stdio servers in the web interface under **Settings → MCP**. It complements the missing features of the built-in `dsh-mcp-client`, including OAuth (authorization code + PKCE, RFC 7591 dynamic client registration) and stdio transport, with static tokens only referencing environment variable names. Tools are registered in the format `mcp__<name>__*`, optional on-demand broker support is available, and workspace isolation is supported. It is licensed under MIT and requires the web profile.
Read MoreReplace DeepSeek Harness's built-in search with MCP search service using dsh-search-mcp
### 正式翻译版本（符合技术文档语境）： Maintained by gxpppp, dsh-search-mcp replaces the built-in `web_search` backend of DeepSeek Harness with search-class MCPs, while keeping the tool name and display unchanged. This article verifies relevant content through the community directory page and GitHub README, including how to switch between Tavily, Brave, Exa, Perplexity and DuckDuckGo on the web settings page, installation commands, key parsing, uninstallation and restoration methods, as well as the fact that the built-in DeepSeek search will be disabled during activation.
Read MoreAdd DeepSeek V4 protocol adaptation for OpenAI client using deepseek-harness
HenryZ838978-maintained deepseek-harness provides protocol-aware adaptation for DeepSeek V4-Pro / V4-Flash, recording 16 protocol features, 10 contracts and 12 probes. This article verifies the access methods including pip, dsh CLI, MCP and SKILL.md, as well as precautions such as reasoning_content 400, prefix caching and the same-name conflict with the official dsh, based on the community directory and GitHub README.
Read More# agents-sdk Skill: Build Stateful AI Agents on Cloudflare
### Official Translation: This article introduces Cloudflare's official agents-sdk Skill: it guides AI assistants to prioritize retrieving Agents documentation, and correctly build stateful Agents on Workers / Durable Objects, covering features such as `setState`, `@callable RPC`, scheduling, Workflows, MCP and React hooks. The article provides installation steps, agents dependency verification, wrangler configuration and a Counter example, and marks precautions such as decorators, migrations and experimental features.
Read MoreUse notion-knowledge-capture to precipitate AI conversations into a Notion knowledge base
### 译文： Introduce OpenAI's curated Agent Skill "notion-knowledge-capture": This tool relies on Notion's official MCP, and organizes decisions and steps from conversations with tools like Cursor, Codex into linkable Notion pages according to templates such as Wiki, How-To, ADR, and FAQ. The article verifies the installation command, the MCP access methods for Codex/Cursor/Claude Code, as well as the official decision capture and How-To sample workflow.
Read MoreBuild ChatGPT Apps with chatgpt-apps Skill: MCP Server + Widget UI
### 正式翻译结果： ### Introduction to OpenAI's official curated Skill chatgpt-apps: How to scaffold, refactor and troubleshoot ChatGPT Apps SDK applications (MCP server + Widget UI) following a docs-first workflow. The article verifies the capabilities of Skills, application prototype classification, upstream example priority, usage boundaries between MCP Apps bridge and window.openai, as well as the installation and activation methods in Codex and Cursor. It also provides prompt examples for paired use with openai-docs and local joint debugging steps for Developer Mode.
Read MoreTurn Notion research into cited structured documents with notion-research-documentation
### Official English Translation: This article introduces the OpenAI curated skill `notion-research-documentation`: under the premise that Notion MCP has been connected, it enables cross-page retrieval and evidence synthesis within Notion, and generates cited documents in templates such as briefings, summaries, comparisons, and full reports. It follows the official SKILL.md instructions for installation and activation (including Codex's `$skill-installer` and Cursor's skills directory), covers the official five-step workflow, competitive product comparison examples, as well as precautions such as permissions, timeliness, and warehouse deprecation.
Read MoreLinear Skill: Let AI Agents Directly Operate Linear Tickets, Not Just Write Code
### 译文： Officially selected by OpenAI: Skill Linear, which enables Agents to read, create and update Issues, projects and team workflows via Linear MCP. This article introduces its positioning, core capabilities, installation steps for Codex/Cursor, typical usage scenarios such as Bug triage and Standup synchronization, as well as precautions including OAuth and rate limits.
Read More# figma-implement-design: Official OpenAI Skill that 1:1 restores Figma design drafts to production-grade code
**figma-implement-design** is an OpenAI-curated Agent Skill that works with the Figma MCP to translate design nodes into deliverable UI code within repositories. This article introduces its seven-step workflow, boundaries with other Skills such as figma-use, prerequisites for the MCP, as well as installation and typical usage in tools like Cursor (/add-plugin figma) and Codex ($skill-installer), helping developers shorten the Design-to-Code iteration cycle.
Read More# MCP-Builder: Official Anthropic Skill to Walk You Through Building High-Quality MCP Servers
MCP-builder is an Agent Skill in Anthropic's official Skills repository, specifically designed to guide developers in creating high-quality MCP servers. Based on the official SKILL.md and reference documents, this article introduces its four-stage workflow (research and planning, implementation, testing, evaluation), dual-stack support for TypeScript/Python, installation and activation methods (Cursor, Claude Code, etc.), typical prompt examples and usage precautions, to help developers systematically encapsulate external APIs into LLM-callable MCP tools.
Read MoreCursor integrates Gmail/Drive/Calendar: Directly read and write Google Workspace within the IDE Agent
On August 3, 2026, Cursor officially released its Google Workspace plugin, allowing coding Agents to access Gmail, Drive and Calendar within the IDE via Google's remote MCP server. This article sorts out the three major plugin capabilities, OAuth installation path, typical scenarios, as well as permissions and security precautions during the preview period, based on the official Changelog, Marketplace and Google developer documentation.
Read MoreGitHub Trending Highlights: Ponytail and Headroom Lead Agent Context Trimming
On August 4, 2026, the top spots on GitHub Trending were taken by Headroom (a tool that compresses 60-95% of Tokens for output/RAG chunks) and Ponytail (a YAGNI decision ladder that reduces code by 54%). This article verifies the official data of the two projects, sorts out two paths for optimizing Agent context costs: input-side compression (Headroom Library/Proxy/MCP) and output-side restraint (Ponytail Claude Code Skill), and provides access suggestions for combined implementation.
Read MoreThe Two AI Coding Powerhouses in 2026: How Claude Code and Cursor Will Divide Labor
In 2026, a new consensus has formed in the developer community: Claude Code (terminal Agent + Claude Agent SDK) and Cursor (AI IDE) are often used in combination rather than being an either-or choice. Based on official documentation from Anthropic and Cursor, as well as public reviews from Faros.ai, FutureProofing.dev and other sources, this paper compares the differences between the two tools in terms of MCP integration, subagents orchestration, multi-model routing, Bugbot review and token efficiency, proposes a practical division of labor scheme such as "Cursor for daily coding, Claude Code for cross-file refactoring and CI automation", and explains that Claude Code can coexist as an extension of Cursor.
Read MoreMajor Revision to MCP Specification: Protocol Layer Fully Switches to Stateless HTTP
On July 28, 2026, the official MCP team released the largest revision since the protocol was launched: the `initialize` handshake and `Mcp-Session-Id` have been removed, and the core of the protocol has been changed to a stateless request/response model. Remote MCP Servers can be deployed behind load balancers like ordinary HTTP services for horizontal scaling; MRTR now supports elicitation interactions in stateless scenarios; gateway-level routing is implemented via the `Mcp-Method`/`Mcp-Name` headers. This article sorts out the changed mechanism, request comparisons, and key points for developer migration.
Read MoreMajor Version Update of MCP: Stateless Core, Official Extensions and Enterprise-level Certification
On July 28, 2026, the Model Context Protocol released its 2026-07-28 specification, marking the largest revision since the protocol's launch. The core changes include: shifting from a bidirectional stateful protocol to a request/response stateless model, removing the initialize handshake and Mcp-Session-Id; upgrading MCP Apps and Tasks to official extensions; and strengthening OAuth 2.0/OIDC authentication to integrate with enterprise IdPs. With monthly SDK downloads exceeding 400 million and over 950 Claude connectors, MCP is becoming the de facto standard for Agent interconnection.
Read MoreStripe's In-House Minions: End-to-End Coding Agent with Over 1300 Weekly PRs and Zero Manual Coding
In February 2026, Stripe officially disclosed that its self-developed coding agent Minions has been deployed at scale: over 1,300 pull requests (PRs) are fully automated by the agent every week, from Slack commands to passing CI checks, with zero manual code writing and only review required for merging. The core architecture includes an isolated Devbox with 10-second warm-up, a custom Harness forked from Block Goose, hybrid orchestration of Blueprint workflows and agents, the Toolshed centralized MCP tool layer (with nearly 500 tools), and a feedback closed loop of up to two rounds of CI. This article sorts out its design highlights and inspirations for enterprise unattended coding agents.
Read MoreModel Context Protocol Releases July 28, 2026 Update: Remove Session, Move Towards Stateless HTTP Deployment
On July 28, 2026, MCP officially released the 2026-07-28 edition specification. The core of the protocol has been changed from a stateless bidirectional model to a stateless request/response model: the initialize handshake and Mcp-Session-Id have been removed, server/discover and Mcp-Method header routing have been added, and list responses support ttlMs caching. SEP-2567 replaces the Session scope with an explicit state handle, and MRTR retains the mid-interaction capability of tools. Extensions such as MCP Apps and Tasks have been formalized, and AWS Bedrock AgentCore, Netlify and others have expressed their support. This article sorts out the key points of the changes and production migration recommendations.
Read MoreMCP's Largest-Ever Update: Stateless Core Shift on 2026-07-28, Agent Infrastructure Goes Enterprise-Grade
On July 28, 2026, the Model Context Protocol released its 2026-07-28 specification: the core of the protocol has shifted from a bidirectional stateful model to a stateless request/response architecture, the initialize handshake and Mcp-Session-Id have been deprecated, and it now supports any instance handling requests after load balancing. It simultaneously introduces MRTR multi-turn interaction, Mcp-Method/Mcp-Name header routing, and list response cache hints, and graduates MCP Apps, Tasks and enterprise managed authorization to official extensions. OAuth hardening includes RFC 9207 iss verification and DCR migration to CIMD. The monthly downloads of Tier 1 SDKs are nearly 500 million, and companies such as Anthropic and Google Cloud have stated that this is a key step toward the productionization of Agents.
Read MoreMCP Server codebase-memory-mcp: Enabling AI Agents to Truly "Remember" Your Codebase
In July 2026, the GitHub Trending hot project codebase-memory-mcp (produced by DeusData) indexes codebases into persistent knowledge graphs via the MCP protocol. It supports parsing 158 programming languages and provides 15 MCP tools, enabling agents such as Claude Code and Cursor to replace file-by-file exploration with structured queries. According to official benchmark tests, the token consumption for structural queries can be reduced by about 120 times, and indexing for Linux kernel-level repositories can be completed in approximately 3 minutes. This article introduces its Hybrid LSP semantic enhancement, core tools, installation method and applicable scenarios.
Read MoreCopilot Code Review Integrates Agent Skills and MCP: Team Guidelines Finally Can Be Included in Reviews
On July 29, 2026, GitHub announced the general availability of Copilot code review support for Agent Skills and MCP servers, which is now available to Pro, Business and Enterprise users. Teams can inject internal coding and review specifications via the SKILL.md file in the .github/skills directory; MCP can pull external contexts such as Jira and document systems in read-only mode, and attribution tags for skills and MCP are added to review comments. This article sorts out the core capabilities of the general release, the configuration examples of SKILL.md, the read-only security boundary of MCP and implementation suggestions.
Read MorexAI Open-Sources Grok Build: Fully Unveils Harness, TUI and Tooling Layer for Rust Coding Agents
On July 15, 2026, SpaceXAI open-sourced Grok Build (the coding Agent behind the grok CLI and its full-screen TUI) under the Apache 2.0 license. The repository xai-org/grok-build quickly gained over 20,000 stars. Based on official announcements and documentation, this article sorts out the Agent loop, tool layer, TUI, and Skills/MCP extension mechanism, introduces ACP integration, local compilation and custom model configuration, and conducts a horizontal comparison with Codex CLI and OpenCode, to help developers understand the architecture and usage of this heavyweight open-source coding Agent Harness.
Read MoreMCP's Largest Version Update: How Stateless Core Enables AI Agents to Truly Enter Production
On July 28, 2026, the Model Context Protocol released its 2026-07-28 specification, marking the largest revision since the protocol launched: the initialize handshake and Mcp-Session-Id have been removed, and the protocol has shifted to a stateless HTTP request/response model, enabling servers to be deployed on Serverless platforms and behind standard load balancers. MCP Apps and Tasks have graduated to official extensions, OAuth 2.0 authorization synchronization has been strengthened, and full support for Claude will be rolled out gradually. This article sorts out six key points including SEP stateless transformation, extension framework, enterprise authorization essentials, and breaking change migration suggestions for production environment evaluation and reference.
Read MorexAI Open-Sources Grok Build: Full Harness of Terminal-Native Coding Agent Exposed
On July 15, 2026, xAI open-sourced the CLI, full-screen TUI and Rust agent runtime of Grok Build coding Agent under the Apache 2.0 license, and disclosed the complete harness implementations including context assembly, tool calling, Skills/Plugins/MCP extensions and so on. This article sorts out the repository architecture layering, extension system and local construction method, and explains the reasons why it has attracted attention as one of the GitHub AI trending repos in July, as well as how developers can make use of this source code.
Read More