The Brev MCP server allows you to manage cloud infrastructure for workspaces and instances across multiple providers (AWS, GCP, Azure, Crusoe, Lambda Labs, Fluidstack, Launchpad, Akash, GCP Alpha).
You can:
Retrieve available instance types for specific cloud providers
Create workspaces by specifying name, instance type, and cloud provider
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., "@Brev train a sentiment analysis model on my dataset"
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.
This is a MCP server implementation for Brev.
The MCP server uses the Brev CLI's API access token and currently set org.
Follow the Brev documentation to download the CLI and login if you haven't already.
If you want to switch your Brev org, run brev set <org-name>
The CLI access token expires every hour. If you have any 403 errors, simply run brev ls to refresh the access token.
Related MCP server: GCP MCP Host on Cloud Run
git clone git@github.com:brevdev/brev-mcp.git
Follow the uv installation guide
On MacOS: ~/Library/Application\ Support/Claude/claude_desktop_config.json
On Windows: %APPDATA%/Claude/claude_desktop_config.json
Add the following to your claude_desktop_config.json:
"mcpServers": {
"brev_mcp": {
"command": "uv",
"args": [
"--directory",
"<path-to-repo>",
"run",
"brev-mcp"
]
}
}To prepare the package for distribution:
Sync dependencies and update lockfile:
uv syncBuild package distributions:
uv buildThis will create source and wheel distributions in the dist/ directory.
Publish to PyPI:
uv publishNote: You'll need to set PyPI credentials via environment variables or command flags:
Token: --token or UV_PUBLISH_TOKEN
Or username/password: --username/UV_PUBLISH_USERNAME and --password/UV_PUBLISH_PASSWORD
Since MCP servers run over stdio, debugging can be challenging. For the best debugging experience, we strongly recommend using the MCP Inspector.
You can launch the MCP Inspector via npm with this command:
npx @modelcontextprotocol/inspector uv --directory /Users/tmontfort/Brev/repos/brev_mcp run brev-mcpUpon launching, the Inspector will display a URL that you can access in your browser to begin debugging.
Create a workspace from an instance type and cloud provider
Inputs: cloud_provider (required), instance_type, name.
Get available instances types for a cloud provider
Inputs: cloud_provider (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 2 tools
The two tools have completely distinct purposes: one creates a workspace and the other retrieves instance types. There is no overlap in functionality, making it impossible for an agent to confuse them.
Both tools follow a consistent verb_noun naming pattern (create_workspace, get_instance_types). The verbs are clear and descriptive, and the style is uniform throughout.
With only two tools, the server feels thin for managing workspaces. It lacks essential operations like listing, updating, or deleting workspaces, which limits its utility for typical lifecycle management.
The toolset is severely incomplete for workspace management. It includes creation and instance type lookup but omits critical operations such as listing existing workspaces, updating configurations, or deleting workspaces, leaving significant gaps in coverage.
Run AI models, create deployments, and manage predictions via cloud API
Discover, run, inspect, build, test, and privately reuse AI workflows.
GPU nodes and clusters, custom experiments, AI training and calculation reports for team workspaces.
AI cloud for agents: launch GPUs, run jobs that stop their own billing, schedule runs.
