MCP (Model Context Protocol) Support

Starting from the version 2.0.0, Goku integrates with the Model Context Protocol (MCP) — which means you can now use Goku programmatically from an LLM agent or any other MCP-aware client.

MCP is an open standard that allows language models and external tools to interoperate through a unified interface: exposing data sources, file systems, APIs or internal logic as “tools” the model can call.

What this enables

• Use Goku from an LLM or AI agent directly — no manual CLI usage required.
• Combine load testing with automated workflows: for instance, trigger a test, gather metrics, and analyze results from within an agent or script.
• Seamless integration into broader toolchains, pipelines or “agentic” workflows, exploiting Goku’s performance-testing features programmatically.

Example usage with an LLM

Once Goku is registered as an MCP tool, you can ask your LLM something like:

"Run a performance test on https://github.com with 2 concurrent users and a total of 30 requests, and provide the 95th percentile response time."

The LLM will translate this into an MCP tool call, run the test through Goku, and return the structured results.

Automatic download (Linux, OSX, WSL)

You can download the latest version of Goku directly to your current directory with the following command:

curl -sSL https://raw.githubusercontent.com/jcaromiq/goku/v2.0.1/scripts/install_mcp.sh | sh

Using Cargo

cargo install goku-mcp

Manual download

Go to the Goku's GitHub Releases page and download the latest .tar.gz file that matches your system. Currently, tarballs are available for the following:

• Linux (x86_64)
• macOS (x86_64)
• Windows (x86_64)

Run MCP

You can use the MCP server

Versioning

CLI is versioned with SemVer v2.0.0.

Contributing

See CONTRIBUTING.md.

Usage manual

Usage: goku [OPTIONS] --target <TARGET>

Options:
  -v, --verbose                      Runs in verbose mode
  -t, --target <TARGET>              URL to be requested using an operation [default: GET] Ex. GET http://localhost:3000/
  -r, --request-body <REQUEST_BODY>  File path for the request body
  -c, --clients <CLIENTS>            Number of concurrent clients [default: 1]
  -i, --iterations <ITERATIONS>      Total number of iterations [default: 1]
  -d, --duration <DURATION>          Duration of the test in second
      --headers <HEADERS>            Headers, multi value in format headerName:HeaderValue
      --scenario <SCENARIO>          Scenario file
      --timeout <timeout_ms>         Timeout value in ms, defaults set to 30000
  -h, --help                         Prints help
  -V, --version                      Prints version information

--target -t

Specifies the operation and url to make the request, default to GET. Format: GET https://localhost:3000

--request-body -r Optional

Specifies the path of file with the body to send.

--clients -c

Specifies the number of concurrent calls to be used, defaults to 1.

--iterations -i

Specifies the total number of calls to be performed, default to 1.

--duration -d

Specifies the duration of the test in seconds.

--headers Optional

Specifies the headers to be sent.

--scenario Optional

Specifies the scenario file in yaml format.

target: POST http://localhost:3000/
clients: 50
requests: 1000
headers:
  - key: "bar"
    value: "foo"
  - key: "Content-Type"
    value: "application/json"

body: "{\"firstName\": \"Terry\",
        \"lastName\": \"Medhurst\",
        \"maidenName\": \"Smitham\",
        \"age\": 50}"

--help

Prints help.

--version

Prints version information.

Simple targets

goku --target "GET http://localhost:3000"
goku --target http://localhost:3000?foo=bar
goku -c 50 -i 1000 --target http://localhost:3000
goku -c 50 --duration 60 --target http://localhost:3000

Targets with custom headers

goku --target "GET http://localhost:3000" --headers Content-Type:application/json --headers bar:foo 

Targets with custom bodies

goku -c 50 -i 1000 -r body.json --target "POST http://localhost:3000"

Targets with custom bodies and headers

goku -r body.json --target "POST http://localhost:3000" --headers Content-Type:application/json --headers bar:foo 

Output

Concurrency level 50
Time taken 4 seconds
Total requests 1000
Mean request time 169.90099999999998 ms
Max request time 415 ms
Min request time 5 ms
95'th percentile: 319 ms
99.9'th percentile: 367 ms

License

See LICENSE.

Donate

If you appreciate all the job done in this project, a small donation is always welcome: