Skip to content
MCP server

Dapr MCP Server

Expose Dapr state, pub/sub, secrets, locks and actors to AI agents as MCP tools

Runs locallyStatus and history
Install
Pricing
Open source
Platforms
self-hosted, CLI, Linux
Runs
Locally
Version
0.0.3
Listing updated
2 days ago

About Dapr MCP Server

Dapr MCP Server is an open-source MCP server for cloud and infrastructure. It runs locally on your machine and is started by your AI client.

Written in Go, this server runs beside a Dapr sidecar and turns the sidecar's building blocks into MCP tools. Because the agent talks to Dapr components rather than to a specific database or broker, swapping Redis for PostgreSQL or Kafka for RabbitMQ needs no change on the agent side. The docs mark the project as alpha.

Tools are registered from the components the sidecar has loaded. Three are always present: get_components, invoke_service and invoke_actor_method. If a deployment has no pub/sub broker, the pub/sub tools are not offered, and hot-reloaded components add or remove tools while clients are notified of the change. The server instructs the model to call get_components first and never invent component names, keys or topics.

It serves stdio for local clients or streamable HTTP with --http for shared deployments; authentication, health endpoints and CORS apply only to HTTP. Authentication can use OIDC, SPIFFE, Dapr Sentry or a hybrid. The docs warn that the server can do anything its sidecar can, including returning secret values to the model, so components and scopes should be limited.

Install

Run in a terminal

claude mcp add dapr-mcp-server --env DAPR_GRPC_PORT='<DAPR_GRPC_PORT>' --env DAPR_API_TOKEN='<DAPR_API_TOKEN>' -- docker run -i --rm -e DAPR_GRPC_PORT -e DAPR_API_TOKEN ghcr.io/dapr/dapr-mcp-server:v0.0.3

Add to claude_desktop_config.json

{
  "mcpServers": {
    "dapr-mcp-server": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "DAPR_GRPC_PORT",
        "-e",
        "DAPR_API_TOKEN",
        "ghcr.io/dapr/dapr-mcp-server:v0.0.3"
      ],
      "env": {
        "DAPR_GRPC_PORT": "<DAPR_GRPC_PORT>",
        "DAPR_API_TOKEN": "<DAPR_API_TOKEN>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "dapr-mcp-server": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "DAPR_GRPC_PORT",
        "-e",
        "DAPR_API_TOKEN",
        "ghcr.io/dapr/dapr-mcp-server:v0.0.3"
      ],
      "env": {
        "DAPR_GRPC_PORT": "<DAPR_GRPC_PORT>",
        "DAPR_API_TOKEN": "<DAPR_API_TOKEN>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "dapr-mcp-server": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "DAPR_GRPC_PORT",
        "-e",
        "DAPR_API_TOKEN",
        "ghcr.io/dapr/dapr-mcp-server:v0.0.3"
      ],
      "env": {
        "DAPR_GRPC_PORT": "<DAPR_GRPC_PORT>",
        "DAPR_API_TOKEN": "<DAPR_API_TOKEN>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "dapr-mcp-server": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "DAPR_GRPC_PORT",
        "-e",
        "DAPR_API_TOKEN",
        "ghcr.io/dapr/dapr-mcp-server:v0.0.3"
      ],
      "env": {
        "DAPR_GRPC_PORT": "<DAPR_GRPC_PORT>",
        "DAPR_API_TOKEN": "<DAPR_API_TOKEN>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "dapr-mcp-server": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "DAPR_GRPC_PORT",
        "-e",
        "DAPR_API_TOKEN",
        "ghcr.io/dapr/dapr-mcp-server:v0.0.3"
      ],
      "env": {
        "DAPR_GRPC_PORT": "<DAPR_GRPC_PORT>",
        "DAPR_API_TOKEN": "<DAPR_API_TOKEN>"
      }
    }
  }
}

Configuration

  • DAPR_GRPC_PORTgRPC port of the Dapr sidecarRequired
  • DAPR_API_TOKENDapr API token, if the sidecar requires one
  • DAPR_MCP_SERVER_LOG_LEVELLog level, written as JSON to stderr

Tools this server provides (9)

Taken from the project documentation; the running server may differ.

get_components
Lists the components loaded in the Dapr sidecar.
invoke_service
Calls another Dapr app through service invocation.
invoke_actor_method
Calls a method on a Dapr actor.
delete_state
Deletes a key from a state store.
execute_transaction
Runs a state transaction.
invoke_output_binding
Calls an output binding.
encrypt_data
Encrypts data with a Dapr crypto component.
get_secret
Reads a secret.
get_bulk_secrets
Reads several secrets.

What it does

Dynamic tool registration
Tools follow the components loaded in the sidecar, with no restart.
Broad building-block coverage
State, pub/sub, bindings, secrets, locks, cryptography, conversation, actors and service invocation.
Stdio and HTTP transports
Run locally for IDE clients or over streamable HTTP for shared setups.
Pluggable authentication
HTTP mode can be gated by OIDC, SPIFFE, Dapr Sentry or a hybrid.
Safety annotations
Each tool carries readOnly, destructive and idempotent hints.
Observability
Tool calls emit OpenTelemetry spans and metrics, with optional log export.

Good for

  • Let an agent read and write through a Dapr state store
  • Publish events or call other Dapr apps from an agent
  • Coordinate workers with a distributed lock
  • Fetch secrets or encrypt data through Dapr components

Pricing

Open source

Prices change. Check docs.dapr.io before you buy.

Dapr MCP Server FAQ

How do I run it?

Start it with dapr run next to a sidecar, or pull the container image ghcr.io/dapr/dapr-mcp-server; the container is meant for Kubernetes with sidecar injection.

Which tools are always available?

get_components, invoke_service and invoke_actor_method.

Which tools should need human confirmation?

The docs list delete_state, execute_transaction, invoke_service, invoke_actor_method, invoke_output_binding and encrypt_data as destructive.

What does the sidecar connection use?

gRPC, with DAPR_GRPC_PORT defaulting to 50001; it retries five times and then exits.

Is it production ready?

The documentation labels it alpha.

Is Dapr MCP Server free?

Yes. Dapr MCP Server is open source, so you can use and self-host it for free.

How do I add Dapr MCP Server to Claude Code?

Run this command in your terminal: claude mcp add dapr-mcp-server --env DAPR_GRPC_PORT='<DAPR_GRPC_PORT>' --env DAPR_API_TOKEN='<DAPR_API_TOKEN>' -- docker run -i --rm -e DAPR_GRPC_PORT -e DAPR_API_TOKEN ghcr.io/dapr/dapr-mcp-server:v0.0.3

What are the best alternatives to Dapr MCP Server?

The closest alternatives to Dapr MCP Server are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Dapr MCP Server

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools