@m4trix/evals 0.26.0 → 0.27.0

package/README.md

@@ -62,7 +62,8 @@ export default defineConfig((): ConfigType => ({
 import { Dataset } from '@m4trix/evals';
 
 export const myDataset = Dataset.define({
-  name: 'My Dataset',
+  name: 'my-dataset',
+  displayName: 'My Dataset',
   includedTags: ['demo'],
 });
 ```
@@ -137,7 +138,7 @@ Group several dataset/evaluator runs under one named config. Each row is either
 `evaluators: [...]` (same module instances discovery loads) or `evaluatorPattern: "..."`
 (wildcard / regex rules from `RunnerApi.resolveEvaluatorsByNamePattern`). Multiple jobs share one `--concurrency` cap.
 
-Optional **`repetitions`** on a row (default `1`) runs each matching test case that many times. Every execution in that group shares the same **`repetitionId`** in the evaluator callback **`meta`**, with **`repetitionIndex`** / **`repetitionCount`**. Evaluator **`meta`** includes **`runConfigName`**: the **`RunConfig`** name (or **`programmatic`** from **`PROGRAMMATIC_RUN_CONFIG`** for API/TUI-only **`runDatasetWith`**). Names may use **kebab-case**, **snake_case**, **camelCase**, etc. (letters, digits, `_`, `-` only, no spaces); resolution is **case-insensitive**.
+Optional **`repetitions`** on a row (default `1`) runs each matching test case that many times. Every execution in that group shares the same **`repetitionId`** in the evaluator callback **`meta`**, with **`repetitionIndex`** / **`repetitionCount`**. Evaluator **`meta`** includes **`datasetName`** (`Dataset.getDisplayLabel()` → `displayName ?? name`) and **`runConfigName`**: the **`RunConfig`** id (or **`programmatic`** from **`PROGRAMMATIC_RUN_CONFIG`** for API/TUI-only **`runDatasetWith`**). **`Dataset`** and **`TestCase`** follow the same naming convention as **`RunConfig`**: **`name`** is the stable id; optional **`displayName`** is unrestricted for UI. Names may use **kebab-case**, **snake_case**, **camelCase**, etc. (letters, digits, `_`, `-` only, no spaces); resolution is **case-insensitive**.
 
 ```ts
 import { RunConfig } from '@m4trix/evals';
@@ -165,7 +166,7 @@ Repeat **`--run-config`** to queue several configs; jobs share one **`--concurre
 
 - `eval-agents`: interactive CLI (starts runs with synthetic meta `programmatic` / `Programmatic`)
 - `eval-agents-simple run --run-config "<RunConfig name>"` (repeatable; case-insensitive match); add **`--ci`** to exit with code **1** if any test case fails
-- `eval-agents-simple generate --dataset "<dataset name>"`
+- `eval-agents-simple generate --dataset "<dataset id>"` (canonical **`Dataset` `name`**, case-insensitive)
 
 ## Default Discovery and Artifacts
 

package/dist/cli-simple.cjs

@@ -55,6 +55,7 @@ function makeEntityIdSchema(brand, label) {
 var RunConfigNameSchema = makeEntityIdSchema("RunConfigName", "RunConfig name");
 makeEntityIdSchema("EvaluatorName", "Evaluator name");
 makeEntityIdSchema("TestCaseName", "Test case name");
+makeEntityIdSchema("DatasetName", "Dataset name");
 function validateWithSchema(schema, raw, context) {
   const trimmed = raw.trim();
   const decode = effect.Schema.decodeUnknownEither(
@@ -611,6 +612,14 @@ function getTestCaseTagList(testCase) {
   return typeof testCase.getTags === "function" ? [...testCase.getTags()] : [];
 }
 
+// src/evals/dataset.ts
+function getDatasetDisplayLabel(dataset) {
+  if (typeof dataset.getDisplayLabel === "function") {
+    return dataset.getDisplayLabel();
+  }
+  return typeof dataset.getName === "function" ? dataset.getName() : "";
+}
+
 // src/evals/metric.ts
 var registry = /* @__PURE__ */ new Map();
 var Metric = {
@@ -1004,7 +1013,7 @@ function processOneEvaluation(task, unit, totalEvaluations, publishEvent, persis
               meta: {
                 triggerId: task.triggerId,
                 runId: evaluatorRunId,
-                datasetId: task.datasetId,
+                datasetName: task.dataset.getDisplayLabel(),
                 repetitionId,
                 repetitionIndex,
                 repetitionCount,
@@ -1419,7 +1428,7 @@ var EffectRunner = class {
       );
       if (!dsCollected) {
         throw new Error(
-          `RunConfig "${rcName}" run[${i}]: dataset "${row.dataset.getName()}" was not found among discovered dataset exports (import the same module instances the scanner loads).`
+          `RunConfig "${rcName}" run[${i}]: dataset "${row.dataset.getDisplayLabel()}" was not found among discovered dataset exports (import the same module instances the scanner loads).`
         );
       }
       let evaluatorIds;
@@ -1554,7 +1563,7 @@ var EffectRunner = class {
     const snapshot = {
       runId,
       datasetId: params.datasetId,
-      datasetName: dataset.dataset.getName(),
+      datasetName: dataset.dataset.getDisplayLabel(),
       evaluatorIds: selectedEvaluators.map((item) => item.id),
       queuedAt: Date.now(),
       totalTestCases: totalEvaluations,
@@ -1575,7 +1584,7 @@ var EffectRunner = class {
       type: "RunQueued",
       runId,
       datasetId: params.datasetId,
-      datasetName: dataset.dataset.getName(),
+      datasetName: dataset.dataset.getDisplayLabel(),
       evaluatorIds: selectedEvaluators.map((item) => item.id),
       totalTestCases: totalEvaluations,
       artifactPath
@@ -1729,7 +1738,7 @@ function getSimpleCliUsage() {
   return [
     "Usage:",
     "  eval-agents-simple run --run-config <name> [--run-config <name> ...] [--concurrency N] [--ci]",
-    "  eval-agents-simple generate --dataset <datasetName>",
+    "  eval-agents-simple generate --dataset <datasetId>",
     "",
     "Options:",
     "  --ci                  With run: exit with code 1 if any test case fails.",
@@ -1797,7 +1806,7 @@ function GenerateView({
       if (!cancelled) {
         setResult({
           count: payload.length,
-          datasetName: dataset.dataset.getName(),
+          datasetName: getDatasetDisplayLabel(dataset.dataset),
           outputPath
         });
         setTimeout(() => onComplete(), 200);
@@ -1858,7 +1867,7 @@ async function generateDatasetJsonCommandPlain(runner, datasetName) {
   const outputPath = createOutputPath(absoluteDatasetPath);
   await promises.writeFile(outputPath, `${JSON.stringify(payload, null, 2)}
 `, "utf8");
-  console.log(`Generated ${payload.length} test cases for dataset "${dataset.dataset.getName()}".`);
+  console.log(`Generated ${payload.length} test cases for dataset "${getDatasetDisplayLabel(dataset.dataset)}".`);
   console.log(`Wrote ${outputPath}`);
 }
 async function generateDatasetJsonCommandInk(runner, datasetName) {
@@ -3109,7 +3118,7 @@ async function main() {
     }
     const genDataset = args.datasetName;
     if (!genDataset) {
-      console.error("Missing required --dataset <datasetName> argument.");
+      console.error("Missing required --dataset <datasetId> argument.");
       printUsageAndExit(1);
     }
     await (useInk ? generateDatasetJsonCommandInk : generateDatasetJsonCommandPlain)(