mcproc

A Model Context Protocol (MCP) server for comfortable background process management on AI agents.

mcproc bridges the gap between AI agent development and traditional command-line workflows. It enables AI agents to manage long-running development processes (like dev servers, build watchers, etc.) while providing developers with full CLI access to monitor and control these same processes.

Simple AI agent-launched processes are stateless and can't manage long-running processes effectively. mcproc solves this by:

- Unified Control: No more confusion about which agent or terminal is running what - all processes are centrally managed
- Context Preservation: Logs are captured and stored, allowing AI agents to debug issues while reviewing logs from earlier
- Developer-Friendly: Full CLI access means you're never locked out of your own development environment

- 🔄Unified Process Management: Start and manage background processes from AI agents via MCP, then monitor them from your terminal
- 👁️Cross-Environment Visibility: Processes started by AI agents are fully accessible via CLI and other agents, and vice versa
- 📝Intelligent Log Management: Capture, persist, and search process logs with powerful regex patterns
- 📁Project-Aware: Automatically groups processes by project context
- 📊Real-time Monitoring: Follow logs in real-time from CLI while AI agents manage the processes
- 🛡️XDG Compliant: Follows XDG Base Directory specification for proper file organization
- ⚡Wait-for-Log: Start processes and wait for specific log patterns to ensure readiness
- 🔍Advanced Search: Time-based filtering, context lines, and regex support for log analysis
- 🧰Toolchain Support: Execute commands through version managers (mise, asdf, nvm, rbenv, etc.)
- 🧹Clean Command: Stop all processes in a project with a single command
- 🌲Process Groups: Automatic cleanup of child processes when stopping parent processes

# Add the tap brew tap neptaco/tap # Install mcproc brew install mcproc

- Rust toolchain (rustc, cargo)
- protobuf compiler:

- macOS:brew install protobuf
- Linux:apt-get install protobuf-compiler

git clone https://github.com/neptaco/mcproc.git cd mcproc cargo build --release # Install to PATH (optional) cargo install --path mcproc

After installing mcproc, you need to register it as an MCP server with your AI assistant.

# Register mcproc as an MCP server claude mcp add mcproc mcproc mcp serve

Configure your MCP client by adding mcproc to your configuration:

{ "mcpServers": { "mcproc": { "command": "mcproc", "args": ["mcp", "serve"] } } }

Once registered, AI agents can use these tools:

- start_process: Start a development server or background process
- stop_process: Stop a running process
- restart_process: Restart a process
- list_processes: List all running processes
- get_process_logs: Retrieve process logs
- search_process_logs: Search through process logs with pattern matching
- get_process_status: Get detailed process information

While AI agents manage processes in the background, you can monitor and control them:

# Start the daemon (if not already running) mcproc daemon start # View all processes (including those started by AI agents) mcproc ps # Follow logs in real-time mcproc logs frontend -f # Multi-process log streaming per project mcproc logs -f # Search through logs mcproc grep backend "error" -C 5 # Stop a process mcproc stop frontend

-

AI agent starts your development server:

Agent: "I'll start the frontend dev server for you" → Uses MCP tool: start_process(name: "frontend", cmd: "npm run dev", wait_for_log: "Server running")

mcproc logs -f # See real-time logs as the server runs

AI agent detects an error and searches logs:

Agent: "Let me check what's causing that error" → Uses MCP tool: search_process_logs(name: "frontend", pattern: "ERROR|WARN", last: "5m")

mcproc grep frontend "ERROR|WARN" -C 3 --last 5m

# Start a process with environment variables mcproc start api --cmd "python app.py" --env PORT=8000 --env DEBUG=true # Wait for a specific log pattern before considering the process ready mcproc start web --cmd "npm run dev" --wait-for-log "Server running on" --wait-timeout 60 # Search logs with time filters mcproc grep api "database.*connection" --since "14:30" --until "15:00" # View logs from multiple processes in the same project mcproc ps mcproc logs web --project myapp -t 100 # Use version managers for Node.js projects mcproc start web --cmd "npm run dev" --toolchain nvm mcproc start api --cmd "yarn start" --toolchain mise # Clean up all processes in a project mcproc clean --project myapp # Force stop all processes in current project mcproc clean --force

mcproc consists of three main components:
- mcproc daemon: A lightweight daemon that manages processes and handles log persistence
- mcproc CLI: Command-line interface for developers to interact with the daemon
- MCP Server: Exposes process management capabilities to AI agents via the Model Context Protocol

- Config:$XDG_CONFIG_HOME/mcproc/config.toml(defaults to~/.config/mcproc/)
- Logs:$XDG_STATE_HOME/mcproc/log/(defaults to~/.local/state/mcproc/log/)
- Runtime:$XDG_RUNTIME_DIR/mcproc/(defaults to/tmp/mcproc-$UID/)

# Clone the repository git clone https://github.com/neptaco/mcproc.git cd mcproc # Build all components cargo build --release # Run tests cargo test # Run with verbose logging RUST_LOG=mcproc=debug cargo run -- daemon start

mcproc/ ├── mcproc/ # CLI and daemon implementation ├── mcp-rs/ # Reusable MCP server library ├── proto/ # Protocol buffer definitions └── docs/ # Architecture and design documentation

Contributions are welcome! Please feel free to submit a Pull Request.

Copyright (c) 2025 Atsuhito Machida (neptaco)

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Create crafted UI components inspired by the best 21st.dev design engineers.

Bring agent evaluations, observability, and synthetic test set generation directly into your IDE for free with Galileo's new MCP server

An MCP server to help AI assistants to answer questions and generate AccelByte Extend SDK code more effectively .

MCP server for AI Diagram Maker — generate beautiful software engineering diagrams directly inside Cursor, Claude Desktop, Claude Code, or any MCP-compatible AI agent

ALAPI MCP Tools,Call hundreds of API interfaces via MCP

AI-powered SVG animation generator that transforms static files into animated SVG components using the Allyson platform

MCP server that gives AI assistants on-demand access to 1,500+ amCharts docs, ~300 code examples, and 1000+ class API references.

APIMatic MCP Server is used to validate OpenAPI specifications using APIMatic. The server processes OpenAPI files and returns validation summaries by leveraging APIMatic’s API.

One shared context layer for AI agents and humans — live API specs, DB schemas, and versioned contracts across repos so every agent and teammate works from the same source of truth.

Build and deploy full-stack Next.js apps with 98 tools for React, AWS, and MongoDB