MCP Server
Dagu includes a built-in Model Context Protocol (MCP) server. There is no separate Dagu MCP package to install: run the Dagu HTTP server and point an MCP client at the /mcp endpoint.
Use MCP when an AI tool should operate a running Dagu server: inspect workflows, read run state, preview workflow edits, apply DAG changes, or control DAG runs through the same authenticated server boundary as the REST API.
What Dagu Exposes
The MCP server is intentionally compact:
| Surface | Purpose |
|---|---|
| Tools | dagu_read, dagu_change, and dagu_execute cover reading state, editing DAG YAML, and controlling runs. |
| Resources | dagu://... resources expose DAG specs, run details, run logs, and built-in MCP references. |
| Prompts | Built-in prompts guide common authoring, editing, and failed-run debugging workflows. |
| Audit events | MCP requests, tool calls, subscriptions, and downstream DAG actions are audit-attributed to the accepted credential. |
The server uses Streamable HTTP. The default local URL is:
http://localhost:8080/mcpIf the Dagu server uses a base path, place /mcp under that base path. For example, a server mounted at /dagu exposes MCP at /dagu/mcp.
Common Flows
| Goal | Start here |
|---|---|
| Connect a client for the first time | Quickstart |
| Configure Codex or Claude Code | Clients |
| Understand the request path and audit flow | Architecture |
| Choose API key, role, surface, and attribution settings | Authentication |
| Read or mutate Dagu through MCP tools | Tools |
Use dagu:// resources or run-completion subscriptions | Resources |
| Review what appears in audit logs | Auditability |
Skill Vs MCP
The Dagu skill and the Dagu MCP server solve related but different problems.
| Integration | Configure | Best for |
|---|---|---|
| Dagu skill | gh skill install dagucloud/dagu dagu | Teaching AI coding tools how to write valid Dagu workflow YAML. |
| Dagu MCP server | http://localhost:8080/mcp | Letting MCP clients read Dagu state, validate or apply DAG edits, and control runs. |
Most AI-assisted workflow authoring setups benefit from both: install the skill for authoring guidance, then connect MCP when the tool should operate a running Dagu server.