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.

Syntax and accessors — Multimeter docs

Each token prefix supports a name made of letters, digits, _, and - (must start with a letter or _). Names are case-sensitive in YAML, but r: and c: names…Markdown (no JavaScript)# Syntax and accessors Each token prefix supports a **name** made of letters, digits, `_`, and `-` (must start with a letter or `_`). Names are case-sensitive in YAML, but `r:` and `c:` names are normalized when looked up (see [Name normalization](#name-normalization)). Supported `r:` and temporal `c:` tokens may append arguments in parentheses. See also: [Dynamic values overview](./index.md) · [Random tokens](./random.md) · [Current tokens](./current.md) ## Environment (`e:`) | Form | Example | Notes | |------|---------|-------| | Angle brackets | `<<e:api_url>>` | **Preferred** inside strings (URLs, header values, body text) | | Plain | `e:api_url` | **Whole scalar only** after `: ` (preserves type). Do not embed bare `e:` in other text — use `<<e:…>>` | | Curly alias | `{{e:api_url}}` | Accepted in YAML outside quotes (rewritten to `<<e:…>>` before parse). Quoted `"{{e:…}}"` stays literal text | `<e:name>` and `e:{name}` are ordinary text. ## Inputs (`i:`) | Form | Example | Notes | |------|---------|-------| | Angle brackets | `<<i:user_id>>` | Use inside strings | | Plain | `i:user_id` | Entire value after `: ` only (not inside arbitrary text like `hi i:user_id`) | | Curly alias | `{{i:user_id}}` | Accepted outside quotes; normalized to bare / `<<i:…>>` on parse | ## Outputs (`o:`) — tests only | Form | Example | Notes | |------|---------|-------| | Angle brackets | `<<o:token>>` | Read `outputs.token` inside strings | | Plain | `o:token` | Standalone value / check expressions | | Set key | `o:token: value` or `o:user.name: value` | Writes `outputs.token` / `outputs.user.name` (only on `set`) | ```yaml outputs: token: null user: null steps: - set: o:token: "abc" o:user.name: "alice" - print: "token=<<o:token>>" - check: <<o:token>> == "abc" ``` Equivalent JS forms `${outputs.token}` and `${outputs.user.name}` remain valid. ## Random (`r:`) and current (`c:`) | Form | Example | Notes | |------|---------|-------| | Angle brackets | `<<r:uuid>>`, `<<c:date>>` | Use inside strings | | Plain | `r:uuid`, `c:epoch` | Entire value after `: ` preserves native type | | Curly alias | `{{r:uuid}}`, `{{c:epoch}}` | Accepted outside quotes (required for whole-value — unquoted `{{…}}` is invalid YAML flow). Normalized to bare / `<<…>>` on parse | Random generators accept documented range/length arguments — see [Random tokens](./random.md). Temporal current tokens accept signed or future/past duration offsets — see [Current tokens](./current.md). ```yaml started_at: c:datetime(-1d2m1s) expires_at: c:utc_datetime(+1h1m) expires_epoch: c:epoch(+30m) ``` Examples mixing prefixes: ```yaml url: <<e:api_url>>/users/<<i:user_id>> headers: X-Req: req-<<r:uuid>> X-Now: <<c:date>> <<c:time>> body: id: r:int created_at: c:epoch active: r:bool username: i:username ``` ## Accessors Append an accessor path after the token name to use part of a string, array, or object value. Supported on all prefixes (`e:`, `i:`, `o:`, `r:`, `c:`). | Accessor | Example | Meaning | |----------|---------|---------| | Property | `<<e:user.name>>` | Object field access | | Index | `<<i:tags[0]>>` | Array index or string character | | Slice | `<<e:token[0:6]>>` | JS `slice(start, end)` — end-exclusive | | Open start | `<<i:message[:4]>>` | From start through index 4 | | Open end | `<<i:message[1:]>>` | From index 1 through end | | Bracket key | `<<e:data[key]>>` | Property when name is a valid identifier | ```yaml headers: Authorization: Bearer <<e:token>> X-Token-Prefix: <<e:token[0:6]>> inputs: username: alice role: admin body: user_initial: <<i:username[0]>> role_short: <<i:role[0:3]>> ``` Input defaults can compose sibling inputs and env vars across multiple resolution passes: ```yaml inputs: card: e:card seq: e:seq short: <<i:card[0:4]>> id: <<i:card>>_<<i:seq>> ``` ## Name normalization For `r:` and `c:` lookups only, names are normalized before matching: - camelCase → snake_case (`firstName` → `first_name`) - hyphens and spaces → underscores (`my-token` → `my_token`) - lowercased (`ALLCAPS` → `allcaps`) So `r:first-name`, `r:first_name`, and `r:firstName` all resolve to the same generator. Environment and input names are **not** normalized — use the exact key from `variables:` / `inputs:`. For AI agents · llms.txt · Home