Multimeter
This MCP server lets an AI assistant work with Multimeter .mmt files across docs, examples, APIs, tests, validation, formatting, and execution.
Read Multimeter YAML documentation by topic and pack (
min/full).List example
.mmtfiles by type/category, optionally with full content.Discover workspace APIs and inspect API files, optionally including content.
Get compact API cards with method, URL, inputs, outputs, and suggested paths.
Scaffold valid smoke or example tests from an existing API
.mmt.Suggest compact
expect/assertpatches from API outputs or JSON response bodies.Validate
.mmtfiles and get structured errors with fix suggestions.Format
.mmtfiles using canonical field ordering and style rules.Run Multimeter test/API files with inputs/env and get structured pass/fail, logs, and errors.
The server explicitly discourages using
testlight,npx,npm, shell, or direct Node server execution.
Download the video (MP4) ยท More demos
๐ Start with a request. Grow into a platform.
Multimeter is Git-native API testing in VS Code. Requests, tests, mocks, and docs are YAML .mmt files in your repo.
Start with a single HTTP request.
Grow into tests, suites, mocks, reports, documentation, and CI when you need them.
All in the same tool. No migration. No second product.
Related MCP server: jmeter-mcp-server
๐ฏ What you get
Simple by default
โ Git-native, file-based YAML
โ Lightweight โ no account, no cloud lock-in
โ Collaboration through pull requests, like code
โ The same files locally and in CI
A full testing platform when you need it
โ HTTP, WebSocket, GraphQL, and gRPC
โ Multi-step flows and test suites
โ Mock servers
โ Generated documentation
โ Reports
โ CI with
testlight
AI in the same files
โ Generate tests from an API or a description (Cursor, Copilot, Claude via MCP)
โ Judge replies โ semantic similarity, or open-ended checks like how funny a response is
โ Bring your own model (Ollama or cloud)
โก Start simple
Multimeter is a VS Code-native extension. All you need:
Click Install button in Multimeter VS Code Extension
Open Get Started from the
Activity barand follow the instructions
There you run a POST request as follows:
type: api
title: Simple POST
url: https://test.mmt.dev/echo
method: post
format: json
body:
message: helloThat's enough.
No suites.
No mocks.
No workflows.
No complexity.
๐งช Need tests?
type: test
steps:
- http: https://test.mmt.dev/echo
method: get
expect:
status: 200Still simple.
Still Git-native.
Still easy to review.
๐ Scale without switching tools
As your project grows, Multimeter grows with it.
Add:
Test suites
Mock servers
Documentation
Workflow execution
Structured reporting
CI artifacts
Only when you need them.
Everything stays in the same ecosystem.
๐ค MCP, AI generation, and AI judges
Cursor, Copilot, and Claude write and run the same .mmt files you edit in VS Code. MCP (mmt-mcp) gives them scaffold_test, validate, format, and run โ generate tests from an API or a description, then keep them in Git.
Judge an API answer in the same test. Compare semantic similarity to an expected reply, or check something open-ended โ for example, how funny the response is. Bring your own model (Ollama or cloud). An alternative to a separate Promptfoo eval stack.
See MCP docs ยท Judge docs
๐ Built for reliable CI
Multimeter validates test definitions before execution.
That means:
โ Earlier feedback
โ More deterministic execution
โ Fewer surprises in CI
โ Easier debugging
โ Reproducible results
GitHub Actions:
- uses: actions/checkout@v6
- uses: mshobeyri/testlight-action@v1
with:
file: tests/suite.mmt
report: junit
report-file: results/junit.xmlOr from a terminal using the Multimeter CLI called testlight.
npm install -g mmt-testlight
testlight run tests/suite.mmtSee the GitHub Action and Testlight install.
๐ฆ Everything stays in Git
Your code, tests, mocks, documentation, reports, and environment settings live in the same repository.
โ Version controlled
โ Code and tests evolve together
โ Reviewable through pull requests
โ Easy to move and share
โ No platform lock-in โ YAML in Git, plus Convert to Postman as an exit path
โ AI can update code and tests together
โ Environment variables never go missing
โ Historical test results stay with the project
๐ง Philosophy
Most API tools focus on requests.
Multimeter focuses on behavior.
Instead of asking:
"Did this request return the expected response?"
Multimeter helps you answer:
"Does this system still behave correctly?"
License
Multimeter is licensed under the Apache License 2.0.
Demos ยท Documentation ยท Website ยท GitHub
Available Tools
9 toolsapi_cardCompact API cardARead-only
Return a small API summary (method, url, inputs, outputs, suggested paths) without dumping the full file. Prefer this over reading the whole .mmt or OpenAPI when generating or planning a test. Next step for a new test: scaffold_test. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: apiPath (required), workspaceRoot (required).
discover_apiDiscover workspace APIsARead-only
Call when listing APIs or inspecting one API before scaffolding a test. For a single API, prefer api_card. For new tests, prefer scaffold_test once apiPath is known. selectedApi omits full file content unless includeContent is true. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: workspaceRoot (required), apiPath, includeContent.
formatFormat .mmt fileBRead-only
Format a Multimeter file using canonical field ordering and style rules. Call after validate passes when normalizing generated or edited YAML. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: file (required), workspaceRoot.
list_examplesList Multimeter examplesARead-only
Call when you need example .mmt structure before creating or modifying a file. Prefer this over searching the repo or guessing YAML layout. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: type, category, maxItems, includeContent.
read_documentationRead Multimeter docsARead-only
Call when you need Multimeter YAML syntax rules. Defaults to pack "min" (small). Use pack "full" only for rare/advanced syntax. For new tests from an API, prefer scaffold_test over loading docs. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: pack, topic.
runRun .mmt fileA
Execute a Multimeter test or API file and return structured pass/fail output, logs, and errors. ONLY supported way to run .mmt files from Copilot โ call this directly when the user asks to run or execute. Do NOT use testlight CLI, npx, npm install, shell scripts, or node dist/mcp/server.js.
Inputs: file (required), env, quiet, inputs, +1 more.
scaffold_testScaffold test from APIARead-only
REQUIRED first step when generating a new Multimeter test from an existing API .mmt. Returns valid smoke (or example) test YAML, suggested path, import alias, and a compact apiCard. Write the yaml, apply only minimal edits, then validate. Do not invent a blank test from scratch. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: apiPath (required), workspaceRoot (required), alias, outPath, +1 more.
suggest_assertionsSuggest assert/expect patchesARead-only
Suggest compact expect/assert YAML patches from API outputs and/or a JSON response body. Use after a run or when tightening a scaffolded smoke test โ patch only, do not rewrite the file. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: body, style, status, stepId, +4 more.
validateValidate .mmt fileARead-only
REQUIRED after every create or modify operation on a .mmt file. Call immediately after editing a file and before telling the user the task is done. Returns structured errors and fix suggestions. Do not use testlight, npx, npm, shell, or node dist/mcp/server.js.
Inputs: file (required), expectedType, workspaceRoot.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelogโฆ
TDQS
Scored across 9 tools
Most tools map to distinct workflow stages: learning syntax, seeing examples, discovering APIs, scaffolding, validating, formatting, and running. discover_api and api_card overlap for single-API inspection, but the descriptions make the preferred choice clear.
Tool names mostly follow an imperative verb_noun pattern such as read_documentation, list_examples, discover_api, scaffold_test, validate, and run. api_card breaks the pattern as a noun-only name, but the inconsistency is minor and not confusing.
Nine tools is well-scoped for this server's purpose of discovering, generating, validating, formatting, and running Multimeter tests. Each tool serves a clear role in the workflow without redundancy or bloat.
The toolset covers the core lifecycle from API discovery and test scaffolding to assertion suggestions, validation, formatting, and execution. A few peripheral capabilities like listing or deleting existing tests are absent, but agents can complete the primary authoring workflow without dead ends.
Maintenance
Related MCP Connectors
3rd Generation Testing (3TG) โ generate deterministic test suites from Markdown spec tables via MCP.
MEOK MCP Test MCP โ golden-file + schema-drift + tool-failure tests for any MCP server. Drop-in
Validate JSON, YAML, XML and CSV with exact line/column errors and silent-corruption warnings.
End-to-end API testing โ generate and run tests from OpenAPI, curl, Postman, or real user traffic.
Related MCP Servers
- AlicenseAqualityCmaintenanceCLI-first API testing tool with MCP server. Tests are .tarn.yaml; failures come back as structured JSON for AI agents to branch on. Tools: tarn_run, tarn_validate, tarn_fix_plan, tarn_inspect, tarn_rerun_failed.145MIT
- AlicenseBqualityDmaintenanceMCP server for generating, validating, running, and reporting Apache JMeter API performance tests from structured API sources.1364 npmMIT
- AlicenseNot gradedqualityCmaintenanceEnables running YAML-defined test suites against MCP servers and agent workflows, with live or simulated execution, assertions, and regression reporting.72 npmMIT
- AlicenseBqualityBmaintenanceEnables AI agents to drive API testing from YAML definitions, executing and chaining requests without hand-written curl, while keeping credentials out of the conversation and responses token-cheap, plus a VANDER-based QA review of endpoints.21MIT