Process Wrapper

This is a mcp server and "track-it" script that conspire to allow claude, gemini, etc to see any logs produced by your binary. The logs are automatically tee'd and their location recorded so the AI can find out where they are and read them. Avoid copy and paste. :-)

python

GitHub Stars

python

Language

7 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "scottmsilver-mcp-track-it": {
      "command": "python3",
      "args": [
        "/path/to/mcp-track-it/mcp_server.py"
      ],
      "env": {
        "MCP_PROCESS_REGISTRY_DB": "/path/to/process_registry.db",
        "MCP_PROCESS_WRAPPER_LOG_DIR": "/path/to/logs"
      }
    }
  }
}

You can track and observe running processes through a dedicated MCP server that exposes a read-only interface for Claude. It captures stdout and stderr independently, stores metadata in SQLite for history, and provides real-time log access and safe process observation without execution rights.

How to use

You will interact with the MCP server through an MCP client. Start a tracked process via the wrapper, then query its status, view logs, or inspect outputs using the MCP interface. The wrapper ensures logs are preserved, processes are cataloged, and Claude can observe without being able to execute commands.

How to install

# Prerequisites
# Ensure Python is installed
python3 --version

# Install the MCP dependency
pip install mcp

# Run the installer script
./install-and-test.sh

This will check your Python installation, test the track-it wrapper, configure Claude's MCP settings, and run a test process to verify everything works.

## Advanced configuration

Customize where logs are written and how the MCP server is wired by setting environment variables and using the provided configuration snippet.

json { "mcpServers": { "process-wrapper": { "command": "python3", "args": ["/path/to/mcp-track-it/mcp_server.py"], "env": { "MCP_PROCESS_WRAPPER_LOG_DIR": "/path/to/logs", "MCP_PROCESS_REGISTRY_DB": "/path/to/process_registry.db" } } } }

## Troubleshooting

If you encounter permission issues, ensure the track-it script is executable. If Claude cannot see processes, restart Claude’s MCP integration after installation. If logs don’t appear, verify write permissions and sufficient disk space in the log directory.

## Notes and tips

The wrapper supports passing environment variables to the tracked process, maintains separate stdout and stderr logs, and stores a persistent history of tracked processes for querying later. Real-time monitoring lets you tail logs as processes run, and Ctrl+C handling ensures clean shutdowns.

## Available tools

### Environment passthrough

Pass environment variables to tracked processes so they can receive configuration from the MCP wrapper.

### Real-time log tailing

Tail stdout and stderr as processes run to provide live visibility to Claude and users.

### Persistent process registry

Store process metadata in SQLite for queryable history across restarts.

### Read-only MCP interface

Claude can observe processes and logs without executing commands.

Built by

VeilStrat

AI signals for GTM teams