Workspace MCP
Integrates with Docker for container management operations.
Integrates with Figma for accessing and manipulating design files.
Provides read-only Git operations including status, log, and diff.
Integrates with GitHub for repository management, issue tracking, and other GitHub API operations.
Integrates with PostgreSQL for database querying and management.
Integrates with Python for scripting, execution, and package management.
Integrates with React for component development and project scaffolding.
Integrates with Rust for compilation, execution, and project management.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Workspace MCPshow my workspaces"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Workspace MCP
Workspace MCP is a generic, secure local development platform that exposes controlled developer capabilities to AI clients through the Model Context Protocol (MCP).
It is not a Merciless Janitors Studios product. MJS is one consumer and may ship an extension package—the core must work for any developer who has never heard of MJS.
Mission
Allow AI assistants to work naturally inside explicitly authorized development workspaces—with deny-by-default security, audit logging, and capability-aware adaptation to each repository.
Related MCP server: DevPilot MCP
Status
Attribute | Value |
Phase | Phase 2 — M2 search + commands implemented |
MCP server | M1+M2 stdio server ( |
Product maturity |
|
OS/capsule | Agentic Delivery OS v1.5.7 |
Initial authorized root |
|
M2 Tools (Implemented)
Tool | Capability | Description |
|
| Search filenames or content |
|
| Run allowlisted commands by ID |
Permission profiles: observer (read+search), maintainer (+commands). See config/examples/commands.json.
M1 Tools (Implemented)
Tool | Capability | Description |
|
| List registered workspaces in grant |
|
| Directory listing |
|
| Read UTF-8 file content |
|
| Git porcelain v2 status |
|
| Bounded one-line log |
|
| Unstaged or staged diff |
|
| Capability detection report |
Writes, search, commands, and extensions are M2+ — see docs/hds/product/roadmap.md.
Installation
cd ~/Documents/dev/workspace-mcp
npm install
npm run buildConfiguration
Copy examples and edit paths for your machine:
mkdir -p ~/.config/workspace-mcp/grants
cp config/examples/workspaces.json ~/.config/workspace-mcp/
cp config/examples/grants/local-dev.json ~/.config/workspace-mcp/grants/
cp config/examples/commands.json ~/.config/workspace-mcp/
chmod 600 ~/.config/workspace-mcp/grants/local-dev.jsonOverride config directory: WORKSPACE_MCP_CONFIG_DIR
Cursor / MCP Client
Add to MCP settings (stdio):
{
"mcpServers": {
"workspace-mcp": {
"command": "node",
"args": ["/absolute/path/to/workspace-mcp/dist/index.js"],
"env": {
"WORKSPACE_MCP_PRINCIPAL_ID": "local-dev"
}
}
}
}Client smoke tests are M5 — configuration above is not yet validated against all clients.
Development
npm test # 15 tests — security + auth + tools
npm run build
npm run dev # stdio server via tsxCore vs Extensions
Layer | Responsibility |
Core | Workspace registration, FS, Git, search, safe commands, audit, permissions, capability detection, extension host |
Extensions | ADOS, hDS, MJS, Universe Engine, Docker, GitHub, Figma, PostgreSQL, Node, Python, Rust, React |
The core must not depend on any extension.
Documentation Map
Topic | Location |
M1 contract |
|
Installation |
|
Tool catalog (M1) |
|
Extension architecture |
|
Security doctrine |
|
Roadmap |
|
Claim Boundaries
Not production-ready · M1 read-only only · Client integrations unverified (M5)
No writes/search/commands until M2/M3
This server cannot be deployed
Maintenance
Related MCP Connectors
Zero-secret MCP gateway for AI agents: risk-scored, audited calls with human-in-the-loop approval.
- gatewayOAuthai.sealgate
MCP gateway with runtime security policy, tool-call-level control, and audit of agent actions.
Security gateway for AI agents: policy, approval, and audited execution, no secrets shared.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Related MCP Servers
- AlicenseBqualityDmaintenanceProduction-grade MCP server that gives AI agents safe access to your local dev environment: filesystem, databases, processes, and OpenAPI specs.1526 npm3MIT
- FlicenseNot gradedqualityBmaintenanceAn AI-powered development assistant that enables LLMs to securely interact with local development tools including Git, filesystem, Docker, PostgreSQL, and GitHub via the Model Context Protocol.-
- AlicenseCqualityAmaintenanceSecure agent coding runtime for local Git repos with policy enforcement, RBAC, sessions, approval workflow, and sandboxed writes, optionally connectable to ChatGPT via Secure MCP Tunnel.85MIT
- AlicenseNot gradedqualityAmaintenanceA secure, local-first MCP server for read-only inspection and troubleshooting of development environments, exposing narrow, typed, auditable capabilities for repository inspection, log summarization, Docker review, and security scanning without granting unrestricted machine access.MIT