mstar-harness
btspoony/mstar-harness
61GitHub starsMulti-agent harness engineering workflows for DeepSeek Harness, with skills, commands, and optional TypeScript-enforced gates.
Discover DSH plugins for permission controls, security auditing, policy enforcement, plugin vetting, credentials, recovery, and safer agent operation.
btspoony/mstar-harness
61GitHub starsMulti-agent harness engineering workflows for DeepSeek Harness, with skills, commands, and optional TypeScript-enforced gates.
jermaine123123/agent-context-editor
61GitHub starsVisual context editor plugin for DeepSeek Harness to search, filter, exclude, and edit conversation turns.
Yan-Zero/dsh-codex
61GitHub starsUse your ChatGPT subscription in DeepSeek Harness through OpenAI's Codex sign-in flow
MichengAI/dsh-skills-manager
60GitHub starsSafely discover, inspect, toggle, import, and manage local agent skills inside DeepSeek Harness Web.
vlln/plugin-registry
58GitHub starsDSH 插件生态基建:薄控制台(浏览器面板管理官方 repository 插件,0 patch)+ make-dsh-plugin skill 官方插件开发引导
AgentDebugX/AgentDebugX
58GitHub stars【EMNLP 2026 Demo】A debugging framework for agentic AI systems: diagnose failures, attribute root causes, recover with evidence, and validate fixes through reruns.
HakureiMonika/dsh-sandbox-escalation-fix
56GitHub starsSession-aware sandbox escalation compatibility plugin for DeepSeek Harness/DSH第三方模型会话沙箱升级兼容插件
linhut/gongwen-skill
56GitHub stars公文全流程处理工具——基于 GB/T 9704《党政机关公文格式》 国家标准,面向公文写作、企事业单位材料编制场景,支持 格式检查与修复、内容优化(Word 原生修订+批注/差异对比版)、模板生成、Markdown 转公文、版头版记页码注入、事实核验、风格增强 等完整能力。原生支持 DeepSeek Harness (DSH) 技能系统,打包为可被 AI Agent 直接调用的 Skill,完全自包含,克隆即用。
opencues/opencues
53GitHub starsA cross-host inline LLM platform that includes a beta DeepSeek Harness integration.
THEWOLFWALKER/dsh-notifier
51GitHub starsA DeepSeek Harness notification plugin with a model-facing notify tool, automatic session-event alerts, and multi-channel delivery.
PerryLink/dsh-fund-research
49GitHub starsDeepSeek Harness tool plugin that produces traceable, versioned Markdown research reports for Chinese public mutual funds.
zhuiyueya/dsh-im-gateway
48GitHub starsAggregate IM gateway plugin that connects DeepSeek Harness agents to 25+ chat platforms.
extracurricular-ai/dsh-filesnap
43GitHub starsRewind DeepSeek Harness conversations and workspace files together without using Git.
PerryLink/dsh-plugin-guide
40GitHub starsDeepSeek Harness plugin-development knowledge base registered as an agent skill, paired with the dsh-plugin-dev CLI toolchain.
xmutfyh/dsh-plugin-writing-guard
40GitHub starsA local, deterministic DSH plugin for auditing academic manuscripts for AI-style patterns, evidence drift, and journal-fit signals.
amlyczz/dsh-lark-link
39GitHub starsA DeepSeek Harness plugin that bridges DSH agents with Feishu/Lark chats.
PlutoKeating/dsh-lark-bot
39GitHub starsBridge DeepSeek Harness into Feishu and Lark with QR scan setup, visual settings, and remote agent execution.
Blank-not-black/dsh-Remote
38GitHub starsMobile and desktop remote console for DeepSeek Harness with token authentication, session control, and file transfers.
jing-hy/picturereader
36GitHub starsDeepSeek Harness image-understanding plugin for text-only models, with local image tools, OCR, document rendering, and an optional external VLM bridge.
PerryLink/dsh-data-quality
35GitHub starsDeterministic data profiling, cleaning, and verification tools and Cordis service for DeepSeek Harness.
PerryLink/dsh-doublecheck
35GitHub starsDelivery quality gate and engineering discipline bundle for DeepSeek Harness.
feiyang-dev/dsh-usage-plugin
34GitHub starsTrack DeepSeek Harness model call tokens, peak/off-peak billing, account balance, and export usage statistics.
fishzjp/qa-skills
33GitHub stars让 AI 像资深测试工程师一样工作:面向 AI Agent 的测试工程 Skill 框架——11 Skills + 共享知识库 + 类型决策矩阵(Claude Code / dsh 等 Agent 可用)
bihangchi9-creator/lark-agent-bridge
33GitHub starsConnects DeepSeek Harness coding agents to Feishu and Lark group chats with isolated per-chat workspaces.
Rianico/dsh-better-edit
33GitHub starsHash-anchored read, edit, and undo tools for DeepSeek Harness that reduce token overhead and prevent line drift.
PensiveFei/dsh-voice-scribe
32GitHub starsDSH voice input plugin: tap Alt to talk, get text in composer. Web Speech default (zero config), optional OpenAI-compatible ASR, polish via DSH LLM.
AcidGr/dsh-web-lan-access
32GitHub starsDeepSeek Harness Web UI plugin that enables LAN access and injects a UUID polyfill for plain-HTTP remote origins.
PGZXB/dsh-feishu
31GitHub starsPanel-driven Feishu / Lark interactive chat and control UI for DeepSeek Harness.
chouxiaohuai/dsh-uiskin-theme
29GitHub starsBlue Glass Theme (ocean fantasy) for DeepSeek Harness Web — static profile bundle. 海洋幻想主题:海洋背景、玻璃气泡、海洋侧边栏、鲸鱼设置按钮、炫彩模型文字。安装: dsh plugin --profile web add github:chouxiaohuai/uiskin-theme
a735624258/dsh-skill-picker
29GitHub starsA DSH Web GUI skill picker that lets users browse and search installed skills, then inserts the native `/skill-name` invocation into the composer.
ChongCyrus/Vibe-Mathematics
28GitHub starsVibe Mathematics —— 多代理数学问题求解与形式化验证框架
Dely0/dsh-personal-workbench
28GitHub starsDSH 个人工作台:日历 + 任务列表 + AI 澄清/拆解/执行/复盘 | Personal workbench for DeepSeek Harness Web: calendar + task list + AI assistant
PerryLink/dsh-claude-move
27GitHub starsApproval-gated migration plugin that imports sessions and context from Claude Code, Codex, OpenCode, and Hermes into DeepSeek Harness.
fb0sh/dsh-pentester
27GitHub stars基于 DeepSeek Harness 的多 Agent PTES 渗透测试编排插件,支持自动化侦察、漏洞分析、验证与报告,使用 Docker/Kali 隔离工具箱 | Multi-agent PTES penetration testing plugin for DeepSeek Harness with automated recon, vulnerability analysis, validation, reporting, and Docker/Kali toolbox
A3Boy/dsh-web-tools
27GitHub starsMulti-provider web search and fetch plugin for DeepSeek Harness supporting 8 search engines and social platforms.
omdsh-dev/dsh-plugin-check
26GitHub starsDSH 插件健康检查工具:扫描插件仓库的清单协议 / patch 格式 / 构建陷阱 / hub 收录状态,零依赖只读,注册 plugin_check 工具
Security and policy plugins inspect, restrict, audit, or recover parts of a DeepSeek Harness environment. This category covers permission rules, plugin vetting, configuration audits, credential controls, access gates, rollback and recovery mechanisms, policy enforcement, and security-focused diagnostics. Protection and enforcement levels differ by implementation, so the Registry keeps source and verification evidence visible instead of reducing plugins to a generic safe label.
When evaluating a security plugin, first determine whether it reports findings or actively blocks and modifies behavior. Review its enforcement point, required privileges, documented fail-open or fail-closed behavior, files and credentials it accesses, dependencies, lifecycle scripts, repository activity, and compatibility evidence. A Security Signal describes observable evidence; it is not itself a security certification.