DeepSeek Harness 沙箱插件

The World's First Unified Virtual Filesystem For AI Agents

Open-source CMA-compatible agent runtime for any model, with MCP tools, sandboxed sessions, audit, replay, and a local console. Includes a native DeepSeek Harness bundle over stdio MCP.

k8e.sh - OpenSource Agentic AI Sandbox Matrix

ANOLISA (Agentic Nexus Operating Layer & Interface System Architecture) | Agentic OS with runtime, security, observability, and Tokenless response compression for lower token usage and cost.

Server-grade gateway that turns DeepSeek Harness into a multi-tenant platform: remote access + auto HTTPS, subuser permissions & quotas, sandbox enforcement, encrypted auth, audit log.

Second-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent returns structured allow/deny verdicts with reasons, fail-closed by default, fully auditable from the session log (approval/asked -> autoReview/verdict -> approval/decided).

原生UI界面。面向 DeepSeek Harness Web 的虚拟机沙箱插件：它基于 OrbStack 为每个会话提供独立的 debian/alpine 沙箱虚拟机，在会话视图环中新增「虚拟机」页签，支持查看/启动/休眠/删除和详细配置展示，同时为模型提供 `vm_list`、`vm_create`、`vm_exec`、`vm_delete` 工具，并内置全局运行上限、闲置自动休眠、归档/删除自动清理等资源治理能力。

DSHBox 是在 Android 上运行 [DeepSeek Harness](DSH）的沙箱 App：Android 原生 UI + PRoot 用户态 Linux 沙箱，DSH WebUI 通过系统浏览器使用，后续可升级为APP原生嵌入。

DeepSeek Harness 自动审批插件：在 workspace-write 沙箱之上叠加确定性规则 + LLM 安全审批，自动模式不放宽沙箱、fail-closed。 Safe auto-approval for DeepSeek Harness — deterministic rules + LLM review on top of the workspace-write sandbox. Auto mode without ever granting full-access.

DeepSeek Harness (DSH) Docker 容器化封装方案，支持安全沙箱、局域网中继与数据持久化。 / Hardened Docker container for DeepSeek Harness (DSH) with LAN access relay and persistence.

dsh-seatbelt 方面沙箱增强尝试使用非exec工具直接对接系统seatbelt相关api

Multi-repo workflow isolation skill for DeepSeek Harness - one isolated entry agent per repo, one controller for cross-project dispatch, model routing & hash-chained ledger

Community plugins for the DeepSeek Harness (dsh): workspace file explorer for the Web UI (Files tab — tree, markdown preview, sandboxed reads)