Work-queue kernel for AI coding agents: dependencies, worktree isolation, quality gates, recovery
Connect Runs on your machine
claude mcp add ddflow-mcp -e DDFLOW_REPO=<DDFLOW_REPO> -e DDFLOW_AGENT=<DDFLOW_AGENT> -- uvx ddflow-mcp==0.2.9{
"mcpServers": {
"ddflow-mcp": {
"command": "uvx",
"args": [
"ddflow-mcp==0.2.9"
],
"env": {
"DDFLOW_REPO": "<DDFLOW_REPO>",
"DDFLOW_AGENT": "<DDFLOW_AGENT>"
}
}
}
}{
"mcpServers": {
"ddflow-mcp": {
"command": "uvx",
"args": [
"ddflow-mcp==0.2.9"
],
"env": {
"DDFLOW_REPO": "<DDFLOW_REPO>",
"DDFLOW_AGENT": "<DDFLOW_AGENT>"
}
}
}
}{
"servers": {
"ddflow-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"ddflow-mcp==0.2.9"
],
"env": {
"DDFLOW_REPO": "<DDFLOW_REPO>",
"DDFLOW_AGENT": "<DDFLOW_AGENT>"
}
}
}
}Replace the values in angle brackets with your own. Keys stay on your machine; we never see them.
Tools
This server runs locally, so its tools are listed by your MCP client once it starts.
Environment variables
| Name | Description |
|---|---|
DDFLOW_REPOOptional | Repository root ddflow manages. Defaults to the client's working directory, resolved to the primary checkout even when called from inside a git worktree. |
DDFLOW_AGENTOptional | Stable identity for this agent. Shards the event log so concurrent agents never write the same file. Defaults to host plus the git worktree's name, so every process in one worktree agrees and two parallel agents differ. A harness running SEVERAL agents inside one worktree must set this, or their wo… |
Details
| Package | [email protected] (PYPI, stdio) |
| Package | docker.io/delian/ddflow-mcp:0.2.9 (OCI, stdio) |
| Package | ghcr.io/delian/ddflow-mcp:0.2.9 (OCI, stdio) |
| First published | 24 Oct 2026 |
| Last updated | 24 Oct 2026 |
| Registry name | io.github.delian/ddflow-mcp |
More developer tools servers See all
Listing data from the official MCP registry. Status and tools from NestDaddy's daily check. Browse all MCP servers.