loading…
Search for a command to run...
loading…
45+ first-class, OS and backend specific MCP tools that let AI coding agents create, manage, read, edit, and write files inside dev containers, keeping your mac
45+ first-class, OS and backend specific MCP tools that let AI coding agents create, manage, read, edit, and write files inside dev containers, keeping your machine clean and your environment reproducible and deploy-friendly. Use local Docker, GitHub Codespaces, Azure, AWS, Google Cloud or Kubernetes seamlessly!
Give your AI agent its own dev environment — not yours.
devcontainer-mcp is an MCP server that lets AI coding agents create, manage, and work inside dev containers across three backends: local Docker, DevPod, and GitHub Codespaces. The agent builds, tests, and ships code in an isolated container — your laptop stays clean.
Works with GitHub Copilot, Claude, Cursor, and any MCP-compatible client.
When AI agents write code, they need to run it somewhere. Today that means your host machine:
The devcontainer spec already defines reproducible, container-based dev environments. Every major project ships a .devcontainer/devcontainer.json. But AI agents can't use them — until now.
devcontainer-mcp exposes 45 MCP tools that let any AI agent:
Agent: "Let me build this project..."
→ auth_status("github") → picks account
→ codespaces_create(auth: "github-you", repo: "your/repo")
→ codespaces_ssh(auth: "github-you", codespace: "...", command: "cargo build")
→ ✅ Built in the cloud. Your laptop did nothing.
curl -fsSL https://raw.githubusercontent.com/aniongithub/devcontainer-mcp/main/install.sh | bash
Invoke-RestMethod https://github.com/aniongithub/devcontainer-mcp/releases/latest/download/install.ps1 | Invoke-Expression
How it works: The binary runs inside WSL; MCP clients on Windows launch it via
wsl ~/.local/bin/devcontainer-mcp serve. The stdio transport works transparently across the WSL boundary. WSL 2 is required — install it withwsl --installif you haven't already.
Backend CLIs (devpod, devcontainer, gh) are detected at runtime — if one is missing, the MCP server returns a helpful error with install instructions.
Binaries available for linux-x64, linux-arm64, darwin-x64, and darwin-arm64.
graph TD
A[AI Agent / MCP Client] -->|stdio JSON-RPC| B[devcontainer-mcp]
subgraph "devcontainer-mcp"
B --> C[33 MCP Tools]
C --> D[Auth Broker]
C --> E[devcontainer-mcp-core]
end
D -->|opaque handles| C
E -->|subprocess| F[DevPod CLI]
E -->|subprocess| G[devcontainer CLI]
E -->|subprocess| H[gh CLI]
E -->|bollard API| I[Docker Engine]
F --> J[Docker / K8s / Cloud VMs]
G --> K[Local Docker]
H --> L[GitHub Codespaces]
| Backend | Best for | Requires | Auth needed? |
|---|---|---|---|
devcontainer CLI (devcontainer_*) |
Local Docker — fast, simple | @devcontainers/cli + Docker | No |
DevPod (devpod_*) |
Multi-cloud: Docker, K8s, AWS, Azure, GCP | DevPod CLI | Optional (cloud providers) |
Codespaces (codespaces_*) |
GitHub-hosted cloud environments | gh CLI | Yes (auth handle) |
The agent never sees raw tokens. Instead:
auth_status(provider) — list available accounts and scopesauth_login(provider, scopes?) — initiate login, opens browser, handles device codesauth_select(id) — switch the active accountauth_logout(id) — revoke credentialsCodespaces tools require an auth handle (e.g. "github-aniongithub"). The MCP server resolves it to the real token on each call via the CLI's native keyring.
Supported providers: GitHub, AWS, Azure, GCP, Kubernetes
| Tool | Description |
|---|---|
auth_status |
Check auth for a provider — returns handles, accounts, scopes |
auth_login |
Initiate login or refresh scopes — browser + device code flow |
auth_select |
Switch the active account for a provider |
auth_logout |
Revoke credentials for an account |
| Tool | Description |
|---|---|
devpod_up |
Create and start a workspace from a git URL, local path, or image |
devpod_stop |
Stop a running workspace |
devpod_delete |
Delete a workspace and its resources |
devpod_build |
Build a workspace image without starting it |
devpod_status |
Get workspace state (Running, Stopped, Busy, NotFound) |
devpod_list |
List all workspaces with IDs, sources, providers, and status |
devpod_ssh |
Execute a command inside a workspace via SSH |
devpod_logs |
Get workspace logs |
devpod_provider_list |
List all configured providers |
devpod_provider_add |
Add a new provider |
devpod_provider_delete |
Remove a provider |
devpod_context_list |
List all contexts |
devpod_context_use |
Switch to a different context |
devpod_container_inspect |
Docker inspect — labels, ports, mounts, state |
devpod_container_logs |
Stream container logs via Docker API |
devpod_file_read |
Read file content with optional line range |
devpod_file_write |
Create or overwrite a file (auto-creates parent dirs) |
devpod_file_edit |
Surgical string replacement — old_str → new_str |
devpod_file_list |
List directory contents (non-hidden, 2 levels deep) |
| Tool | Description |
|---|---|
devcontainer_up |
Create and start a local dev container |
devcontainer_exec |
Execute a command inside a running dev container |
devcontainer_build |
Build a dev container image |
devcontainer_read_config |
Read merged devcontainer configuration as JSON |
devcontainer_stop |
Stop a dev container (via Docker API) |
devcontainer_remove |
Remove a dev container and its resources |
devcontainer_status |
Get dev container state by workspace folder |
devcontainer_file_read |
Read file content with optional line range |
devcontainer_file_write |
Create or overwrite a file (auto-creates parent dirs) |
devcontainer_file_edit |
Surgical string replacement — old_str → new_str |
devcontainer_file_list |
List directory contents (non-hidden, 2 levels deep) |
auth handle| Tool | Description |
|---|---|
codespaces_create |
Create a new codespace for a repository |
codespaces_list |
List your codespaces with state and machine info |
codespaces_ssh |
Execute a command inside a codespace via SSH |
codespaces_stop |
Stop a running codespace |
codespaces_delete |
Delete a codespace |
codespaces_view |
View detailed codespace info (state, machine, config) |
codespaces_ports |
List forwarded ports with visibility and URLs |
codespaces_file_read |
Read file content with optional line range |
codespaces_file_write |
Create or overwrite a file (auto-creates parent dirs) |
codespaces_file_edit |
Surgical string replacement — old_str → new_str |
codespaces_file_list |
List directory contents (non-hidden, 2 levels deep) |
{
"mcpServers": {
"devcontainer-mcp": {
"command": "devcontainer-mcp",
"args": ["serve"]
}
}
}
{
"mcpServers": {
"devcontainer-mcp": {
"command": "wsl",
"args": ["~/.local/bin/devcontainer-mcp", "serve"]
}
}
}
Install backend CLIs as needed — the MCP server detects them at runtime and returns helpful errors if missing:
npm install -g @devcontainers/cli + Dockerauth_login toolWhen devcontainer_up, devpod_up, or codespaces_create fails, the full build output (including errors) is returned to the agent. The agent can read the error, fix the Dockerfile or devcontainer.json, and retry — making the dev environment a dynamic, agent-managed asset rather than a static prerequisite.
This project eats its own dogfood — development happens inside its own devcontainer.
# Using the devcontainer CLI
devcontainer up --workspace-folder .
devcontainer exec --workspace-folder . cargo build --workspace
devcontainer exec --workspace-folder . cargo test --workspace
devcontainer exec --workspace-folder . cargo build --release -p devcontainer-mcp
# Or using DevPod
devpod up . --id devcontainer-mcp --provider docker --open-ide=false
devpod ssh devcontainer-mcp --command "cd /workspaces/devcontainer-mcp && cargo build --workspace"
cargo check, cargo test, cargo clippy, cargo fmt run automaticallyВыполни в терминале:
claude mcp add devcontainer-mcp -- npx