Multimeter is a Git-native REST client and API testing tool for VS Code. Requests, tests, mocks, and docs are YAML .mmt files in Git — the same files in the visual editor and in CI via testlight. Apache 2.0. No account.

How do I use Multimeter with Cursor, Copilot, or Claude?

Install the VS Code/Cursor extension or add MCP with npx -y mmt-mcp. Agents should call scaffold_test, validate, format, and run instead of guessing .mmt syntax.

Yes. Apache License 2.0. No account required.

HTTP file — Multimeter docs

Use .http and .https files with Multimeter without giving up REST Client or JetBrains HTTP Client workflows.Markdown (no JavaScript)# HTTP file Use `.http` and `.https` files with Multimeter without giving up REST Client or JetBrains HTTP Client workflows. ## Run directly See [Spec editor](./spec-editor.md) for the {{btn:list-tree}} request picker, **All** vs single request, and {{btn:save-as}}. In VS Code, open a `.http` or `.https` file with **Open With...** → **Multimeter HTTP Test Editor**, or **Open as MMT**. The left editor stays on the HTTP source. ## Import in tests Reference HTTP files from `type: test` `.mmt` files. Multimeter converts the file to a test flow internally: ```yaml type: test title: Reuse HTTP requests import: posts: echo_posts.http steps: - call: posts ``` See [import](../files/test/import.md). ## Convert to MMT Right-click a `.http`, `.https`, or `.rest` file in the Explorer and choose **Convert to MMT...**. Multimeter generates `api/` and `tests/` files — one API per request, plus a test file with sequential `call` steps for multi-request files. Pick the generated files, a destination folder, and a collision policy (skip, overwrite, or rename). Example output: [HTTP convert example](../../examples/intermediate/25_convert_to_mmt/http/README.md) Full sandbox HTTP file: [professional HTTP convert](../../examples/professional/04_convert_to_mmt/http/README.md) ## Supported syntax Targets VS Code REST Client and JetBrains HTTP Client syntax: - Request separators (`###`), names (`# @name`), `METHOD URL` lines (with optional `HTTP/1.1` / `HTTP/2`) - Headers and raw bodies (JSON, XML, text detection) - File variables (`@host = ...`) and `{{variable}}` references - System variables (`{{$guid}}`, `{{$uuid}}`, `{{$randomInt}}`, `{{$timestamp}}`, `{{$datetime}}`) - Request chaining (`{{login.response.body.$.token}}`) - Basic status checks from response scripts (`response.status === 200`) Each request block maps to an internal `http` step with `debug: true` so responses appear in the run panel. ## Limitations Not fully supported yet: pre-request scripts, full response handler execution, multipart/form-data parts, body file includes (`< ./payload.json`), cookie jars, redirect control, digest auth, AWS SigV4, and other tool-specific auth helpers. Use editable `.mmt` files for advanced flow control, data-driven loops, mocks, load tests, and full environment presets. ## See also - [HTTP file examples](../../examples/intermediate/15_http_files/README.md) — start with `auth_user_flow.http` and pick **All** or one request - [Bruno](./bruno.md) · [Postman](./postman.md) · [Test import](../files/test/import.md) For AI agents · llms.txt · Home