Perfscale MCP
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Perfscale MCPrun a load test on my API with 100 virtual users for 5 minutes"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
perfscale MCP servers
Model Context Protocol servers for perfscale:
Package | Server | Use case |
| Local: run k6/locust/native load tests, lint and manage test/config YAML via the OSS | |
| Cloud: machines, tests, runs, metrics, and audit data from perfscale.su / perfscale.ru |
Quick start
OSS (local)
Requires the perfscale CLI on PATH (or PERFSCALE_BIN).
{
"mcpServers": {
"perfscale": {
"command": "npx",
"args": ["-y", "@perfscale/mcp"]
}
}
}Tools: run_test, lint, get_schema, parse_summary, list_actions,
list_configs, read_config, write_test, write_config, update_config,
remove_config.
Controlplane (cloud)
Create an API token in the dashboard (Settings → API Tokens; Scale plan: 1 per user, Enterprise: 5 per user), then:
{
"mcpServers": {
"perfscale-cloud": {
"command": "npx",
"args": ["-y", "@perfscale/controlplane-mcp"],
"env": {
"PERFSCALE_API_URL": "https://perfscale.su",
"PERFSCALE_API_TOKEN": "psk_..."
}
}
}
}Tools: whoami, limits, usage, list_machines, get_machine,
list_tests, get_test, write_test, list_configs, write_config,
list_runs, get_run, get_run_logs, runs_by_machine, run_test,
metrics_catalog, query_metrics, get_otel_timeseries, list_dashboards,
list_git_repos, sync_git_repo, list_env_vars (values masked),
audit_log.
Every tool call is metered against the workspace's monthly AI tool-call
budget — usage shows where you stand, limits shows the plan.
Composite mode (workspace MCP servers)
By default the server also aggregates the remote MCP servers attached to
your workspace (dashboard → Settings → AI → MCP Servers): their tools
appear next to the core ones, prefixed with the server's name (grafana →
grafana_search_dashboards), and calls are proxied through. Only remote
transports are supported (streamable HTTP, SSE fallback) — never local
commands. Resolving the attached servers requires an admin token: header
values are secrets, so a regular member's token gets the core tools only. A
remote that is down is skipped with a warning; the core surface always
works. Set PERFSCALE_MCP_COMPOSITE=off to force the plain core server.
Related MCP server: k6 MCP Server
Development
pnpm install
pnpm test # vitest across all packages
pnpm build # tsc -b per packageReleasing
Bump the version in the package's package.json, then tag and push:
git tag v0.2.0 && git push origin v0.2.0CI publishes to npm (release.yml); packages whose version is already on
the registry are skipped, so packages version independently. Requires the
NPM_TOKEN repo secret.
This server cannot be deployed
Maintenance
Related MCP Connectors
Drive OctoPerf load testing from any AI agent: import, edit, validate, run scenarios, read metrics.
Load & browser performance testing — drive MaxoPerf from your AI agent with your API key.
Load testing and synthetic monitoring platform: test with Playwright, Browser Bot, or Protocol Bots.
Discover Playwright workflows, start runs, and inspect results in Playrunner Cloud.
Related MCP Servers
AlicenseNot gradedqualityBmaintenanceEnables AI agents to manage performance testing workflows on BlazeMeter's cloud platform through natural language interactions.26Apache 2.0- FlicenseNot gradedqualityDmaintenanceEnables running k6 load tests with customizable duration and virtual users via natural language, with real-time output and LLM-powered analysis.1-
- AlicenseNot gradedqualityDmaintenanceEnables AI models (Claude, ChatGPT, GitHub Copilot) to run and analyze local tests, rerun failures, and orchestrate QA workflows using existing UI and API test frameworks.6 npmMIT
- AlicenseNot gradedqualityBmaintenanceConverts non-functional requirements and code into secure, traceable JMeter performance tests with coverage analysis, plan generation, execution, and GitHub draft PRs.173 npmMIT