AIDevTools - Sidekick MCP Server
LocalAbout AIDevTools - Sidekick MCP Server
AIDevTools is a set of MCP servers and tools for AI-assisted software development. Its Sidekick server combines process control, agent-to-agent Q&A, and notifications, while a separate proxy bridges stdio clients to SSE servers.
Key features
- 01Start, monitor, and control long-running processes
- 02Stream process output with managed buffers
- 03Agent specialist Q&A and communication
- 04Audio notifications on macOS
- 05Proxy between stdio and SSE MCP
Use cases
- 01Track and control background development processes
- 02Route questions between agents and specialists
- 03Receive task audio notifications on macOS
- 04Connect a stdio client to an SSE MCP server
MCP support
Sidekick exposes process management, agent communication, and notification tools over MCP. The suite also includes stdio2sse, which proxies requests between stdio MCP clients and SSE MCP servers.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEAIDevTools
MCP servers and tools for AI-powered development.
Components
🚀 Sidekick
Process management, agent communication, and notification MCP server for AI agents.
Features:
- Spawn and manage long-running processes
- Real-time output streaming with ring buffers
- Interactive process control (stdin/stdout/stderr)
- Audio notifications (macOS)
- Agent Q&A system for specialist communication
- TUI mode for visual process monitoring
- Cross-platform: Linux, macOS, Windows
Install:
curl -sSL https://raw.githubusercontent.com/eliezedeck/AIDevTools/main/sidekick/install.sh | bash
Usage:
# TUI mode (default)
sidekick
# Stdio mode for Claude Desktop
sidekick --sse=false
# SSE server mode with custom port
sidekick --port 6060
# Add to Claude Desktop (stdio mode)
claude mcp add sidekick ~/.local/bin/sidekick --args "--sse=false"
🌉 stdio2sse
Proxy between stdio-based MCP clients and SSE-based MCP servers.
Features:
- Connect Claude Desktop to SSE-only MCP servers
- Automatic tool discovery and proxying
- Transparent request/response forwarding
- Async architecture for reliable communication
Install:
curl -sSL https://raw.githubusercontent.com/eliezedeck/AIDevTools/main/stdio2sse/install.sh | bash
Usage:
# Bridge to an SSE server
stdio2sse --sse-url http://localhost:5050/sse
# Add to Claude Desktop
claude mcp add my-sse-server ~/.local/bin/stdio2sse --args "--sse-url" "http://localhost:5050/sse"
Requirements
- Go 1.23+ (for building from source)
- Claude Desktop or other MCP-compatible clients
Building from Source
git clone https://github.com/eliezedeck/AIDevTools.git
cd AIDevTools
# Build sidekick
cd sidekick && go build
# Build stdiobridge
cd ../stdiobridge && go build
API Reference
Sidekick Tools
Process Management:
spawn_process- Start a new process with options (delay, buffer size, environment)spawn_multiple_processes- Launch multiple processes sequentiallyget_partial_process_output- Get incremental output (tail -f functionality)get_full_process_output- Get all output in memorysend_process_input- Send stdin input to a running processlist_processes- List all tracked processes and their statuskill_process- Terminate a tracked processget_process_status- Get detailed process information
Agent Communication:
get_next_question- Register as a specialist and wait for questionsanswer_question- Provide an answer to a received questionask_specialist- Ask a question to a specialist agentget_answer- Retrieve answer for a previously asked questionlist_specialists- List all available specialist agents
Notifications (macOS only for now):
notifications_speak- Play sound and speak text (max 50 words)
License
MIT License - see LICENSE file for details.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.