
Artillery Mcp
Drive Artillery 2.x load tests, Fargate runs, and project-based scenarios from your coding agent instead of hand-rolling CLI commands.
Overview
Artillery MCP is a MCP server for the Ship phase that runs Artillery 2.x load tests with full flags, Fargate, parsers, and project launchers from your agent.
What is this MCP server?
- Wraps Artillery 2.x with full CLI flag coverage from MCP tools
- Supports AWS Fargate distributed runs for heavier load scenarios
- Raw-text parser for interpreting Artillery output inside the agent
- Project launcher with optional .artillery-mcp.config.json for run_project_lt and counterBreakdown
- Configurable ARTILLERY_TIMEOUT_MS (default 30 minutes) and Artillery Cloud recording via API key
- Server version 0.1.7 (npm and OCI docker.io/kosiakmd/artillery-mcp:0.1.7)
- Default ARTILLERY_TIMEOUT_MS 1800000 (30 minutes)
- stdio transport; MIT-licensed Artillery 2.x MCP wrapper
What problem does it solve?
You know you need load tests before launch but bouncing between Artillery docs, shell flags, and cloud setup eats solo-builder time.
Who is it for?
Indie builders with an Artillery project who want agent-driven perf checks before shipping or after infra changes.
Skip if: Teams with no HTTP/API surface to test or who already run fully automated CI load suites without agent involvement.
What do I get? / Deliverables
Your agent can start Artillery runs, optional Fargate jobs, and structured project tests from one registered MCP server with sane timeouts and cloud recording.
- Agent-triggered Artillery load test executions with parsed output
- Optional distributed Fargate test runs
- Project-scenario runs when .artillery-mcp.config.json is configured
Recommended MCP Servers
Journey fit
How it compares
MCP load-testing integration for Artillery CLI, not a generic unit-test or security-audit skill.
Common Questions / FAQ
Who is Artillery MCP for?
Solo and small-team builders using Claude Code or Cursor who ship APIs or web apps and already use or want Artillery 2.x for performance testing.
When should I use Artillery MCP?
Use it in the ship phase when you need to run or tune load tests, spin Fargate runs, or launch project scenarios before release or after scaling changes.
How do I add Artillery MCP to my agent?
Register the stdio server via npm @kosiakmd/artillery-mcp or the Docker OCI image, set ARTILLERY_WORKDIR to your repo root, and optionally ARTILLERY_MCP_CONFIG and ARTILLERY_CLOUD_API_KEY.