# maxoperf
**Category:** ๐ป Developer Tools
**Repository:** https://github.com/maxoperf/mcp
**Views:** 0
**Installs:** 0
**Upvotes:** 0
**Directory Page:** https://allmcps.com/mcp/maxoperf
## Description
Load & browser performance testing โ drive MaxoPerf from your AI agent with your API key.
## Claude Desktop Quick Installation
Heuristic fallback โ verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):
```json
"mcpServers": {
"maxoperf": {
"command": "npx",
"args": ["-y","maxoperf"]
}
}
```
## Documentation & README
# โก MaxoPerf MCP Server
### Drive the entire MaxoPerf platform from any AI agent โ with your API key.
**A remote, hosted [Model Context Protocol](https://modelcontextprotocol.io) server that turns Claude, Cursor, Codex, VS Code Copilot, and ChatGPT into a full performance-testing operator. Create tests, launch load runs on managed cloud runners, read results, and get root-cause analysis โ all in natural language, all authenticated with your own key.**
[](https://modelcontextprotocol.io)
[](https://claude.ai)
[](https://cursor.com)
[](https://chatgpt.com)
```
https://app.maxoperf.com/mcp
```
---
## Configuration
It's a **remote, hosted** server โ nothing to install or run. Add this to your MCP client config and set `MAXOPERF_API_KEY` (create one in the console โ **Settings โ API keys**):
```json
{
"mcpServers": {
"maxoperf": {
"type": "http",
"url": "https://app.maxoperf.com/mcp",
"headers": {
"Authorization": "Bearer ${MAXOPERF_API_KEY}"
}
}
}
}
```
Your key is validated on every call by the real platform (auth, tenancy, OpenFGA, audit) โ the server stores nothing and adds no new trust boundary. Revoke the key, access dies instantly. Send the header `X-MaxoPerf-MCP-Mode: read-only` for a look-but-don't-touch session (write tools hidden).
## Install (pick your client)
**Claude Code** โ the plugin bundles the server **and** the agent skill:
```
/plugin marketplace add MaxoPerf/mcp
/plugin install maxoperf
```
Or add just the connector:
```
claude mcp add --transport http maxoperf https://app.maxoperf.com/mcp \
--header "Authorization: Bearer ${MAXOPERF_API_KEY}"
```
**Cursor** ยท **VS Code / Copilot** ยท **Codex** ยท **ChatGPT** ยท **Claude Desktop** โ one-click deeplinks and copy-paste config in [`packaging/`](./packaging). All use the same `mcpServers` block above.
---
## Tools
29 curated tools. Reads default to `response_format: "concise"` (pass `"detailed"` for the full payload); write tools require a non-read-only session, and `cancel_run` is hidden in read-only mode.
**Context & tenancy**
- `whoami` โ Resolve the account + default workspace behind your API key
- `list_workspaces` โ List workspaces visible to the account
- `set_active_workspace` โ Set the active workspace for the session
- `list_projects` โ List projects (with edit/delete permissions)
- `create_project` โ Create a project
**Tests**
- `list_tests` โ List tests (filter by project / workspace / type)
- `get_test` โ Get one test + its validation summary
- `create_test` โ Create a test shell (choose the engine/executor)
- `get_test_overview` โ Run-history overview for a test
**Test files**
- `upload_test_file` โ Upload a script/data file in one call (real 3-step presigned flow)
- `list_test_files` โ List a test's files + upload state
- `download_test_file` โ Get a short-lived download URL for a file
**Runs**
- `start_run` โ Launch a load/browser run on managed cloud runners (idempotent)
- `get_run_status` โ Poll lifecycle status (queued โ running โ passed/failed/cancelled)
- `list_runs` โ Paginated run history with filters
- `cancel_run` โ Cancel a run (destructive; hidden in read-only)
- `rerun_run` โ Re-run from a snapshot or the current test
- `add_runners` โ Scale a live run up at existing locations
**Results**
- `get_run_results` โ KPI overview: throughput, latency percentiles, error rate
- `query_run_metrics` โ Time-series metrics (latency / throughput / errors / load / health)
- `get_run_errors` โ Grouped error rows (message / count / code)
**Diagnostics โ root-cause & anomaly detection**
- `get_run_summary` โ Executive summary + which failure criteria tripped
- `get_run_error_bodies` โ Sampled error request/response bodies + status codes
- `get_run_logs` โ Error-level engine/system log lines
- `get_runner_health` โ Runner CPU/mem trend + targetVus vs peakAchievedVus (vuShortfallPct)
- `detect_run_anomalies` โ Deterministic robust-outlier scan (median/MAD); terminal-gated, low false-positive
**Escape hatch & discovery**
- `call_platform_api` โ Reach any public `/v1/*` endpoint (secrets, environments, schedules, BYOC); admin/internal deny-listed, SSRF-safe
- `get_openapi` โ The public OpenAPI document
- `search_endpoints` โ Keyword search over the API to find the right endpoint
## Prompts
Text recipes that encode the correct tool sequence โ great for chat-only clients that can't read a repo:
- `run-baseline-load-test` โ Start a baseline run and watch it to completion
- `diagnose-latency-regression` โ Compare p95 across two runs
- `summarize-run` โ Plain-language summary of one run
- `plan-and-build-test` โ Turn a goal into project โ test โ upload โ run
- `choose-executor` โ Recommend an engine (k6 / JMeter / Playwright / Selenium)
- `scan-endpoints-for-hotspots` โ Rank likely hotspots from an OpenAPI spec or pasted list
- `setup-secrets-and-envs` โ Wire workspace secrets + multi-env before a run
- `diagnose-run-failure` โ Ranked root cause for a failed run
- `explain-run-anomalies` โ Explain each detected outlier
## Resources
- `maxoperf://openapi` โ The public OpenAPI spec
- `maxoperf://run/{id}` โ A run report summary
---
## Try it
- "Load test https://api.example.com/checkout with 500 users for 5 minutes and fail it if p95 goes over 800ms."
- "Scan my repo for the endpoints most worth load-testing, then build and run a test for the riskiest one."
- "Why did run run-0000000001 fail? Check the errors, the logs, and whether the runners actually reached the target load."
## Pair it with the brain
The MCP server is the **hands**. The bundled **[MaxoPerf agent skill](https://github.com/MaxoPerf/agent-skill)** (`npx @maxoperf/agent-skill install`, included in the Claude plugin, or in [`agent-skill/`](./agent-skill) here) is the **brain** โ it reads your code, finds the hotspots, builds and runs the test, and diagnoses why it broke, driving these tools automatically.
## What's in this repo
| Path | What |
| --- | --- |
| [`.claude-plugin/`](./.claude-plugin) | Claude Code plugin (bundles the MCP connector + the skill) |
| [`agent-skill/`](./agent-skill) | A copy of the `maxoperf` agent skill (canonical home: [MaxoPerf/agent-skill](https://github.com/MaxoPerf/agent-skill)) |
| [`packaging/`](./packaging) | MCP Registry `server.json`, `.mcpb`, VS Code / Cursor deeplinks, Codex / ChatGPT setup |
| [`LAUNCHGUIDE.md`](./LAUNCHGUIDE.md) | MCP directory listing metadata |
---
**[Get started โ](https://maxoperf.com)** ยท **[Agent skill โ](https://github.com/MaxoPerf/agent-skill)** ยท **[Docs โ](https://maxoperf.com/docs)**
Read-only mirror of the MaxoPerf monorepo. File issues at [maxoperf.com](https://maxoperf.com).