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.
95 results
Local-first, self-hosted AI agent runtime with Claude Managed Agents-style APIs, sandboxed sessions, memory, tools, audit, replay, and a local Console.
Deployment scripts for MisakaNet Workers
可弹出侧边栏 · A DeepSeek Harness sidebar that shows artifacts and pops out into a larger web tab.
Design I/O for coding agents: controllable UI generation via declarations (spec/domain/craft/design/components/template) and contracts (skill/evaluator). Use for product UI—console, dashboard, agent-ops, CJK-first apps.
SSH ops and database tools for DSH Web UI: connect to remote servers via SSH terminal/SFTP/tunnels, and connect to MySQL/PostgreSQL/Redis/MongoDB databases. Host half manages sessions with ssh2; client half renders the terminal and database panel.
Hot-pluggable Hakurei Shrine (Touhou Project) skin for the dsh web GUI: shrine day/night backdrops, Reimu figures, framed sidebar & composer, transparent paper UI
dsh web desktop pet: a Niulai calf that lives in the corner, hops around, and moos "mama" when an agent task completes (client-only plugin)
A loop dock for DeepSeek Harness: one harness, multiple agent loops.
DeepSeek Harness plugin: a sidebar "today's usage" card that pops a token usage dashboard (today / 7 / 30 days, 30-day per-model stacked chart, model ranking with usage share, conversation stats). Data comes from the session corpus only (idempotent rebuild, no double counting), persisted across restarts.
Two byte-preserved original whale animation loops for DeepSeek Harness Web.
Probabilistic background Shadow agent reviews for DeepSeek Harness
DSH Web GUI paperclip button: a single 📎 button in the composer that opens a file picker and drops file content into the textarea.
CloudQ integration for DeepSeek Harness with secure credential, workspace, and plugin-management surfaces
SVN (Subversion) tools + sidebar UI for DeepSeek Harness: full command coverage (status/info/diff/log/blame/list/cat/add/delete/mkdir/copy/move/merge/switch/checkout/cleanup/resolve/props/commit...) with UTF-8 Chinese commit logs, plus an SVN panel in dsh-better-sidebar.
Agent run-rhythm guard for DeepSeek Harness: reasoning dead-loop interception (guard) + automatic continuation after premature stops (continue)
企业 AI 资源统一管理平台 —— 基于 DeepSeek Harness「一切皆插件」架构
Native Windows toast + system tray for DeepSeek Harness (dsh): the only dsh plugin with a tray icon. Pops the moment the agent stops running — finished / aborted / error / output limit / waiting for your choice / session closed — body labeled with workspace · session, so you can switch windows during long tasks and glance at the tray instead of watching a spinner. Pure host-side, zero dependencies, zero build.
DeepSeek Harness plugin: content-aware PDF reading tools for vision models (pdf_scan / pdf_read_page / pdf_render_region). Detects figures (vector+raster), tables, math, and two-column layout per page, then routes text pages to extraction and figure/table/math pages to high-DPI region crops rendered for read_image — avoiding DeepSeek's ~800x800 whole-page token ceiling. Backed by PyMuPDF; returns a clear warning (with the exact install command) when Python or a dependency is missing.
Audit deepseek-harness (DSH) plugins before install, guard them at runtime. Static verdict + pre-install audit protocol; supply-chain checks (typosquat, OSV); exfiltration & ransomware detection, honeypot canaries, integrity baseline. Alarm-only; blocks confirmed destructive ops.
Git / source-control plugins for DeepSeek Harness: a local-git provider, a hosted-platform registry (GitHub/GitLab/Bitbucket/Azure DevOps/Gitea), and grouped model-facing tools.
DeepSeek Harness SSH 远程运维插件,支持远程命令、文件管理、JumpServer 与 keyboard-interactive 认证
DSH 容器管理插件:七工具覆盖容器与镜像管理,走官方 subprocess 服务、argv 无 shell 注入,exec 与破坏性 manage 默认审批。
DSH WebUI 控制台:一键检查更新、一键执行更新(核心 npm + 插件 pnpm)、一键重启。
A green interrupt button beside the DeepSeek Harness send box: silently stops a running agent, which then summarizes its progress and asks for your new requirements; the interrupt prompt is customizable.