SkillsMay 12, 2026·2 min read

CoWork OS — Local-First Agent Workspace

CoWork OS is a local-first agent workspace for code, docs, web pages, and automations, with an Electron UI and built-in support for many MCP connectors.

AI Open Source
AI Open Source · Community
Agent ready

Review-first install path

This asset needs a review step. The copied prompt tells the agent to dry-run, show the writes, then proceed only after confirmation.

Needs Confirmation · 66/100Policy: confirm
Agent surface
Any MCP/CLI agent
Kind
Skill
Install
Single
Trust
Trust: Established
Entrypoint
Asset
Review-first command
npx -y tokrepo@latest install ed99140d-3c1d-5649-bb8b-52f01464fbd7 --target codex

Dry-run first, confirm the writes, then run this command.

Intro

CoWork OS is a local-first agent workspace for code, docs, web pages, and automations, with an Electron UI and built-in support for many MCP connectors.

  • Best for: People who want one local app for agentic coding + docs + browser work
  • Works with: Node.js + npm; Electron; optional provider keys (Claude/OpenAI/Gemini/OpenRouter) and MCP connectors (per docs)
  • Setup time: 20–60 minutes

Practical Notes

  • GitHub: 291 stars · 18 forks; pushed 2026-05-12 (verified via GitHub API).
  • Docs docs/getting-started.md uses npm run setup then npm run dev to launch the Electron app + hot reload.
  • README advertises dozens of provider options and many MCP connectors for a single “agent workspace” surface.

Main

Adoption tips:

  1. Start with one workspace and define conventions (branching, lint, CI) so the agent’s edits stay consistent.
  2. Enable only a small set of connectors first; too many tools makes planning noisy and harder to audit.
  3. Use explicit approvals for any action that sends messages, writes to production, or spends money.
  4. Capture “what worked” as repeatable shortcuts and keep them versioned like code.

Local-first is a feature: it keeps files and context near the work, with fewer external dependencies.

FAQ

Q: Is it cloud-based? A: No—README frames it as local-first; your workspace lives on your machine.

Q: Do I need a single LLM provider? A: No—docs list multiple providers and optional failover chains.

Q: How do I keep tool use safe? A: Enable connectors gradually and require approvals for outbound or high-impact actions.

🙏

Source & Thanks

Source: https://github.com/CoWork-OS/CoWork-OS > License: MIT > GitHub stars: 291 · forks: 49

Discussion

Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.

Related Assets

Local Deep Research — Privacy-First AI Research Agent

A self-hosted deep research agent that achieves near-perfect accuracy on benchmarks using local or cloud LLMs, with support for 10+ search engines and fully encrypted processing.

Skills
Script Depot

Eigent — Local Cowork Desktop for AI Agents

Eigent is an open-source cowork desktop for running multi-agent workflows locally. Use it to delegate research, planning, and execution with replays.

Skills
AI Open Source

Colanode — Open-Source Local-First Slack and Notion Alternative

A self-hosted collaboration platform combining real-time team chat, a rich document editor, and a knowledge base in a single local-first application that keeps all data on your own infrastructure.

Skills
AI Open Source

Refact — Local-First AI Coding Assistant

Refact is an open-source, local-first AI coding assistant: install the IDE plugin, run local refact-lsp, and connect a model provider.

Skills
AI Open Source
◈Home🔍Search👤Me