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.

Logging — Multimeter docs

Multimeter writes structured run logs to the VS Code Output panel when you run APIs, tests, or suites from the editor. Use it to inspect inputs and outputs,…Markdown (no JavaScript)# Logging Multimeter writes structured run logs to the VS Code **Output** panel when you run APIs, tests, or suites from the editor. Use it to inspect inputs and outputs, timings, suite exports, and failures without leaving VS Code. ![Multimeter logs in the VS Code Output panel with the Multimeter channel selected and log level filter open](../screenshots/logging-output.png) ## Where logs appear Open **View → Output**, choose **Multimeter** from the channel dropdown, and scroll through the run log. When `multimeter.showLogOnRun` is enabled (default), the Output panel opens automatically on editor runs. | Entry point | Destination | |---|---| | Editor (Send, Run test, Run suite) | Output panel → **Multimeter** channel | | AI assistant (`@multimeter /run`) | Output panel → **Multimeter** channel | | CLI (`testlight`) | Terminal stdout | See [Where logs appear](./logging/where-logs-appear.md) for CLI options such as `--quiet` and `--log-level`. ## Log levels The Output panel includes a built-in level filter (gear icon in the panel header). Available levels: | Filter | Shows | |---|---| | **Trace** | Lowest detail — network request/response summaries during test runs | | **Debug** | Request/response for API runs, environment variables, imported check passes | | **Info** | Inputs, outputs, run lifecycle messages, direct-run check passes (default) | | **Warning** | Non-fatal issues (e.g. example not found) | | **Error** | Check/assert failures, runtime exceptions | | **Off** | Hide all Multimeter log output | Multimeter maps its internal levels (`trace`, `debug`, `info`, `warn`, `error`) to these Output panel levels. A run is marked **failed** if any `error`-level message is logged. For a full breakdown of what each level includes, see [Log levels](./logging/log-levels.md). ## What gets logged Logs are append-only and grouped by run. Typical content: - **API runs** — `Inputs` and `Outputs` sections, finish message with duration (e.g. `API "Get user" finished in 141 ms successfully`); full request/response at `debug` - **Test runs** — check/assert results, `print` output, network details at `trace` when using `call` steps - **Suite runs** — per-item start/finish, aggregated child logs, cancellation warnings - **Suite exports** — paths where report files were written (e.g. JUnit XML) Timestamps and level tags prefix each line (`[info]`, `[debug]`, etc.). ## Tips - Start at **Info** for everyday runs; switch to **Debug** or **Trace** when you need request/response bodies or network details - Use **Error** to focus on failures during large suite runs - The webview **Log** tab inside the test Report panel shows step-level check output during a run; the Output panel keeps the full, persistent log — see [Test logging](./logging/test-logging.md) --- ## See also - [API logging](./logging/api-logging.md) · [Test logging](./logging/test-logging.md) · [Suite logging](./logging/suite-logging.md) - [Testlight](./testlight/index.md) — `--log-level` and `--quiet` for terminal runs - [History](../panels/history.md) — recent requests and responses in the bottom panel (separate from run logs) For AI agents · llms.txt · Home