walkerOS Flow — Claude Code MCP server: what it does, install
This MCP server lets an agent inspect walkerOS event flows, run simulations, create bundles, and send them onward. It runs on the local machine.
claude mcp add --env WALKEROS_TOKEN=<WALKEROS_TOKEN> --env WALKEROS_PROJECT_ID=<WALKEROS_PROJECT_ID> --env WALKEROS_APP_URL=<WALKEROS_APP_URL> --transport stdio mcp -- npx -y @walkeros/mcpWhat walkerOS Flow does
walkerOS Flow is a local MCP process that communicates with clients over stdio. Once connected, an agent can inspect event-flow definitions, test their behavior in a simulation, create a bundle, and push the result. The client starts the server through npx using the @walkeros/mcp package.
Claude Code can register it with WALKEROS_TOKEN. Other setup variants pass WALKEROS_PROJECT_ID and WALKEROS_APP_URL as environment variables as well. The token is required for the connection, so a setup without WALKEROS_TOKEN will not work.
The package handles flows on the local machine. Data held by the walkerOS app is covered by a separate hosted endpoint described in the source, so this local process is a different route to the service.
What it helps with
- inspect walkerOS event flows
- run a flow simulation
- create a deployable bundle
- push a completed flow
Example prompts for the agent
- Inspect this walkerOS event flow and report configuration problems.
- Run a simulation of the flow and explain which events continue through it.
- Create a bundle for this flow and push the result using the connected MCP server.
How to install walkerOS Flow
Platform CLI
claude mcp add --env WALKEROS_TOKEN=<WALKEROS_TOKEN> --env WALKEROS_PROJECT_ID=<WALKEROS_PROJECT_ID> --env WALKEROS_APP_URL=<WALKEROS_APP_URL> --transport stdio mcp -- npx -y @walkeros/mcpReplace <…> with your own values
Platform CLI
codex mcp add mcp --env WALKEROS_TOKEN=<WALKEROS_TOKEN> --env WALKEROS_PROJECT_ID=<WALKEROS_PROJECT_ID> --env WALKEROS_APP_URL=<WALKEROS_APP_URL> -- npx -y @walkeros/mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"mcp": {
"command": "npx",
"args": [
"-y",
"@walkeros/mcp"
],
"env": {
"WALKEROS_TOKEN": "<WALKEROS_TOKEN>",
"WALKEROS_PROJECT_ID": "<WALKEROS_PROJECT_ID>",
"WALKEROS_APP_URL": "<WALKEROS_APP_URL>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@walkeros/mcp"
],
"env": {
"WALKEROS_TOKEN": "<WALKEROS_TOKEN>",
"WALKEROS_PROJECT_ID": "<WALKEROS_PROJECT_ID>",
"WALKEROS_APP_URL": "<WALKEROS_APP_URL>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e WALKEROS_TOKEN=<WALKEROS_TOKEN> -e WALKEROS_PROJECT_ID=<WALKEROS_PROJECT_ID> -e WALKEROS_APP_URL=<WALKEROS_APP_URL> mcp npx -y @walkeros/mcpReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about walkerOS Flow
How do I install walkerOS Flow in Claude Code?
Run claude mcp add flow -e WALKEROS_TOKEN=<WALKEROS_TOKEN> -- npx -y @walkeros/mcp. This registers the MCP server with the token in Claude Code.
What does walkerOS Flow do?
It gives an agent access to walkerOS event flows for inspection, simulation, bundling, and pushing results.
Which environment variables are used by the server?
WALKEROS_TOKEN is used for the basic connection. Set WALKEROS_PROJECT_ID and WALKEROS_APP_URL as well when the client configuration needs a project and app endpoint.
