Outflow MCP runs on your own machine — the client starts it, so there is no endpoint to ping. 34 installs a week from npm. Last commit 17 Jul 2026.
MCP server exposing an Outflow workspace's live architecture graph as context for AI coding agents
Today is the operative word: we check Outflow MCP every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add outflow-mcp -e 'OUTFLOW_API_KEY=<OUTFLOW_API_KEY>' -e 'OUTFLOW_WORKSPACE_ID=<OUTFLOW_WORKSPACE_ID>' -- npx -y outflow-mcp
{
"mcpServers": {
"outflow-mcp": {
"command": "npx",
"args": [
"-y",
"outflow-mcp"
],
"env": {
"OUTFLOW_API_KEY": "<OUTFLOW_API_KEY>",
"OUTFLOW_WORKSPACE_ID": "<OUTFLOW_WORKSPACE_ID>"
}
}
}
}
[mcp_servers.outflow-mcp]
command = "npx"
args = ["-y", "outflow-mcp"]
env = { OUTFLOW_API_KEY = "<OUTFLOW_API_KEY>", OUTFLOW_WORKSPACE_ID = "<OUTFLOW_WORKSPACE_ID>" }
{
"mcpServers": {
"outflow-mcp": {
"command": "npx",
"args": [
"-y",
"outflow-mcp"
],
"env": {
"OUTFLOW_API_KEY": "<OUTFLOW_API_KEY>",
"OUTFLOW_WORKSPACE_ID": "<OUTFLOW_WORKSPACE_ID>"
}
}
}
}
{
"servers": {
"outflow-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"outflow-mcp"
],
"env": {
"OUTFLOW_API_KEY": "<OUTFLOW_API_KEY>",
"OUTFLOW_WORKSPACE_ID": "<OUTFLOW_WORKSPACE_ID>"
}
}
}
}
Put your own values in place of
<OUTFLOW_API_KEY> (API key for the workspace to expose. Create one from workspace settings in Outflow (Settings → API Keys) — the raw key (ofk_...) is shown once), <OUTFLOW_WORKSPACE_ID> (ID of the workspace to expose. Found in the workspace URL (.../workspace/<id>/...) or workspace settings).
The author’s registry entry marks them as required.
Optional settings from the registry entry, left out of the block above:
OUTFLOW_BASE_URL (Base URL of your running Outflow instance. Defaults to http://localhost:3000 if omitted).
MCP server exposing CAST Imaging software architecture insights to AI agents
Semantic code graph MCP server for coding agents
App Store Connect MCP Server for AI coding agents
MCP server that provides architecture design expertise to AI coding agents
Local-first architecture guardrails and project context for JS/TS AI coding agents.
Code knowledge graph over MCP: token-budgeted context for AI coding agents.
On-demand, local code-context MCP server for AI coding agents with Tree-sitter AST.
Context-aware code retrieval MCP server — ranked, gated results for AI coding agents
Answers built from our own checks of this server.