Uniprof

Provides a unified profiling workflow to profile various runtimes using MCP clients and visualize the results.

javascript

395

GitHub Stars

javascript

Language

4 months ago

First Indexed

2 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": {
    "indragiek-uniprof": {
      "command": "npx",
      "args": [
        "-y",
        "uniprof",
        "mcp",
        "run"
      ]
    }
  }
}

You can profile any application without changing its code, using a unified MCP server workflow. This setup lets you run profiling tools in isolated containers or locally, with automatic runtime detection and consistent analysis results that help you identify hotspots across languages and runtimes.

How to use

You use an MCP client to profile and analyze applications in a single step or through a sequence of recorded profiles. Start a profile for your language runtime, then inspect the consolidated results or flamegraphs in your browser. You can also run a recording, analyze the data for hotspots, and visualize flamegraphs for quick performance insights.

How to install

Prerequisites you need before installing and using the MCP server: Node.js and npm or npx, and Docker if you plan to run profilers inside containers. The MCP server is installed globally on your system to enable MCP workflows.

npm install -g uniprof

Install the MCP server integrations that enable automatic MCP client setup using the built-in installer.

uniprof mcp install claudecode
uniprof mcp install cursor
uniprof mcp install vscode

Additional notes

If automatic MCP client installation is not supported for your environment, you can still run an MCP server using a standard stdio transport by executing a command that launches the MCP runner directly.

npx -y uniprof mcp run

Configuration and usage patterns

Profile a program in one step and immediately analyze results. You can also record profiles to a file, then run analysis and visualization later.

# Profile in a single step for multiple runtimes
uniprof python app.py
uniprof node server.js

# Save a profile and analyze later
uniprof record -o profile.json -- python app.py
uniprof analyze profile.json
uniprof visualize profile.json

Available tools

record

Capture a profile from a running program and save it to a file for later analysis.

analyze

Process a saved profile to identify hotspots and generate aggregated insights.

visualize

Open or render flamegraphs and performance visuals in a browser to explore profiling results.

bootstrap

Optional environment check to prepare the MCP runtime for profiling.