managed-agents
Local-first, self-hosted AI agent runtime with Claude Managed Agents-style APIs, sandboxed sessions, memory, tools, audit, replay, and a local Console.
89 results
Local-first, self-hosted AI agent runtime with Claude Managed Agents-style APIs, sandboxed sessions, memory, tools, audit, replay, and a local Console.
DSH inline visualization plugin: a visualize tool + bundled skill let the model render interactive HTML fragments as sandboxed cards in the conversation (Codex /visualize semantics).
Sandbox-first automatic permission policy for DeepSeek Harness
DeepSeek Harness plugin: installs the 'minimal-gitbash' agent preset — the Windows variant of the shipped minimal preset that routes the bash tool through Git for Windows bash (MSYS), with automatic shell discovery and sandbox-aware gating.
dsh-passwords: a server-grade gateway that turns DeepSeek Harness into a multi-tenant platform — remote access + automatic HTTPS, per-subuser permissions & quotas, sandbox enforcement, first-run setup, SQLite auth with at-rest encryption, rate-limit and audit log (bilingual zh/en UI)
Windows-first bash tools for DeepSeek Harness: git_bash (Git Bash) and wsl_bash (WSL2) model tools with bwrap sandbox, approval mode, terminal-style UI cards, and background-job support
Net Access permission preset for DeepSeek Harness (Windows): HTTPS works inside the sandbox while file writes stay confined like workspace-write.
Tensorlake sandbox execution providers for DeepSeek Harness
DSH sandbox extension: configurable trusted writable roots outside the workspace, for both sandboxed CLI commands and the write/edit tools. 为 DSH 默认沙箱增加可配置的「沙箱授权目录」,允许工作区外的受信修改。
Fail-closed microsandbox microVM provider and model-facing guest tools
DeepSeek Harness 的 Docker 容器沙箱部署级插件(与 dsh-plugin-vm-sandbox 全能力对齐):39 个 docker_* 模型工具、快照/回滚、文件传输、端口转发、后台任务、审计、共享/配额/回收、网络策略、多容器并行执行与华丽原生 UI。无需 OrbStack,Linux/Windows/macOS 均可使用。
DSH tool-call compatibility: strips sandbox_permissions/justification from tool-call arguments for third-party models (GPT etc.), and lets the user skip possibly-stuck tool calls with an LLM-facing notice.
让DSH使用你的魔法:把沙箱里的 agent 接上你本地的 HTTP 代理,一条命令抓取被墙的网站、文档和 API(net_fetch 代理隧道抓取 + net_proxy_status 代理诊断)。Give DSH your magic: plug sandboxed agents into your local HTTP proxy and fetch blocked sites, docs and APIs in one shot (net_fetch + net_proxy_status).
Run the DeepSeek Harness agent's filesystem, bash and PTY terminals inside a Docker container, so execution state survives a harness restart.
DSH compatibility plugin that treats redundant non-escalating sandbox permission fields as a no-op
dsh-yolo-mode —— DeepSeek Harness 双面包插件:当会话处于可写沙箱模式且审批策略为 ask 时,用大模型自动裁决沙箱升权申请,支持内置预设与自定义权限层级,并提供宿主 settings + 自发布设置桥(/yolo-mode)与 Web 客户端 UI。
Self-evolving plugin for DeepSeek Harness: observes agent behavior, proposes improvements to its own skills/policies via LLM, validates them in a sandboxed trial agent, and applies versioned mutations with rollback.
DSH 测试沙盒插件:把开发/测试任务路由到进程级隔离的沙盒实例(独立进程 + 独立 DSH_HOME + 独立端口),提供创建/注入/运行/健康检查/停止/销毁/清理与门禁式合回等原生工具,全程零影响本体(3080)。
DSH context-window relief: knowledge base (ctx_index/search), routing enforcement (deny flood tools), sandboxed execution (ctx_execute/batch Think-in-Code), and session continuity (post-compaction restore).
nono (Landlock/Seatbelt) backend for the DeepSeek Harness sandbox seam with a vendored fail-closed executor
OrbStack VM sandbox v0.4.0: 在 v0.3(安全/可靠性)基础上新增 A1 交互终端、A3 场景化一键创建、A4 删除撤销+完成通知、A5 配额一眼可见、A6 协作分组、B1 安全基线 vm_harden、B2 密钥库 vm_secret、C1 增强指标+阈值告警 vm_alert、D1 vm_scp/vm_logs/vm_env/vm_withdraw、D2 调度 UI+用量报表 vm_report、D3 导出分片/远端、D4 累计配额+创建排队。
QuickJS/WASM-isolated WorkflowEngine for running model-written DeepSeek Harness orchestration with bounded resource controls.
DeepSeek Harness 自动审批插件:在 workspace-write 沙箱之上增加 Auto 档——确定性规则放行/拒绝,模糊操作由 LLM 裁决,危险操作转人工。保留工作区沙箱边界,不放宽为 full-access。
A middle permission tier for DeepSeek Harness between workspace-write and danger-full-access: auto-approves harmless commands and operations targeting configured trusted areas, beyond the current workspace.