@dalmia/calibrate-mcp 0.0.92 → 0.0.94

package/bin/mcp-server.js

@@ -53308,9 +53308,9 @@ var init_config = __esm(() => {
   SDK_METADATA = {
     language: "typescript",
     openapiDocVersion: "0.1.0",
-    sdkVersion: "0.0.92",
+    sdkVersion: "0.0.94",
     genVersion: "2.915.1",
-    userAgent: "speakeasy-sdk/mcp-typescript 0.0.92 2.915.1 0.1.0 @dalmia/calibrate-mcp"
+    userAgent: "speakeasy-sdk/mcp-typescript 0.0.94 2.915.1 0.1.0 @dalmia/calibrate-mcp"
   };
 });
 
@@ -58646,7 +58646,7 @@ var init_annotationTasksCreateEvaluatorRun2 = __esm(() => {
 });
 
 // src/models/createjobsrequest.ts
-var ReasoningMode$zodSchema, CreateJobsRequest$zodSchema;
+var ReasoningMode$zodSchema, EvaluatorReasoningModes$zodSchema, CreateJobsRequest$zodSchema;
 var init_createjobsrequest = __esm(() => {
   init_zod();
   ReasoningMode$zodSchema = _enum([
@@ -58654,10 +58654,16 @@ var init_createjobsrequest = __esm(() => {
     "required",
     "hidden"
   ]).describe("How the labelling form treats the reasoning box on each judgement. `optional` shows it, `required` shows it and asks the annotator to fill it in, `hidden` leaves it out");
+  EvaluatorReasoningModes$zodSchema = _enum([
+    "optional",
+    "required",
+    "hidden"
+  ]);
   CreateJobsRequest$zodSchema = object({
     annotator_ids: array(string2()).describe("Annotator IDs to assign, creating one labelling job for each annotator. Must be in your workspace"),
     comments_enabled: boolean2().default(true).describe("When `true`, the labelling form lets the annotator leave a comment on each item"),
     evaluator_ids: array(string2()).nullable().optional().describe("Subset of the task's linked evaluators to show in these jobs. Must be a subset of the current links, an empty list gives a 400. Applies to every annotator's job. Omit (`None`) to snapshot every linked evaluator"),
+    evaluator_reasoning_modes: record(string2(), EvaluatorReasoningModes$zodSchema).nullable().optional().describe("Map of evaluator ID to how the labelling form treats the reasoning box for that evaluator, with the same values as `reasoning_mode`. Every ID must be an evaluator on these jobs, otherwise you get a 400. An evaluator left out uses `reasoning_mode`"),
     item_ids: array(string2()).optional().describe("Item IDs to assign. **Required when `select_all=false`**. Ignored when `select_all=true`"),
     q: string2().nullable().optional().describe("Case-insensitive substring filter on `payload.name`. Applies only when `select_all=true`"),
     reasoning_mode: ReasoningMode$zodSchema.default("optional").describe("How the labelling form treats the reasoning box on each judgement. `optional` shows it, `required` shows it and asks the annotator to fill it in, `hidden` leaves it out"),
@@ -61194,10 +61200,10 @@ var OutputScaleEntry$zodSchema;
 var init_outputscaleentry = __esm(() => {
   init_zod();
   OutputScaleEntry$zodSchema = object({
-    color: string2().nullable().optional().describe("Color to show for this level. Omit for the default"),
-    description: string2().nullable().optional().describe("Rubric text for this level, added to the judge prompt. Omit to leave this level undescribed"),
-    name: string2().describe("Short label for this scale point"),
-    value: any().optional().describe("The value for this scale point. Use a boolean for a `binary` evaluator, a number for a `rating` one")
+    color: string2().nullable().optional().describe("Color to show for this row. Omit for the default"),
+    description: string2().nullable().optional().describe("Rubric text for this row, added to the judge prompt. Omit to leave this row undescribed"),
+    name: string2().describe("Short label for this row"),
+    value: any().optional().describe("The value for this row. Use a boolean for a `binary` evaluator, a number for a `rating` one. For a rating evaluator, give the label for the worst result the lowest number and the label for the best result the highest number")
   }).describe("One entry in an evaluator's output_config.scale. `value` is bool|number|string\ndepending on output_type.");
 });
 
@@ -61207,7 +61213,7 @@ var init_outputconfig = __esm(() => {
   init_zod();
   init_outputscaleentry();
   OutputConfig$zodSchema = object({
-    scale: array(OutputScaleEntry$zodSchema).nullable().optional().describe("The ordered scale points that make up the rubric, each with its label")
+    scale: array(OutputScaleEntry$zodSchema).nullable().optional().describe("The rows of the rating scale, each a number and its label. For a rating evaluator, give the label for the worst result the lowest number and the label for the best result the highest number")
   });
 });
 
@@ -61219,7 +61225,7 @@ var init_evaluatorversioncreate = __esm(() => {
   init_variablespec();
   EvaluatorVersionCreate$zodSchema = object({
     judge_model: string2().describe("The model that runs the judge, named the way its provider does, for example `openai/gpt-4.1` or `anthropic/claude-sonnet-4`"),
-    output_config: OutputConfig$zodSchema.nullable().optional().describe("The scale points and their labels. Required for a `rating` evaluator. A `binary` evaluator uses the default Correct/Wrong labels unless you set your own"),
+    output_config: OutputConfig$zodSchema.nullable().optional().describe("The rows of the rating scale, each a number and its label. Required for a `rating` evaluator. For a rating evaluator, give the label for the worst result the lowest number and the label for the best result the highest number. A `binary` evaluator uses the default Correct/Wrong labels unless you set your own"),
     system_prompt: string2().describe('Judge system prompt. May contain `{{"{{"}}variable}}` placeholders'),
     variables: array(VariableSpec$zodSchema).nullable().optional().describe('Declared prompt variables. Omit if the prompt has no `{{"{{"}}placeholders}}`')
   }).describe("One version of an evaluator: its judge prompt, model, variables, and rubric.");
@@ -61404,7 +61410,7 @@ var init_evaluatorversioncreaterequest = __esm(() => {
   EvaluatorVersionCreateRequest$zodSchema = object({
     judge_model: string2().describe("The model that runs the judge, named the way its provider does, for example `openai/gpt-4.1` or `anthropic/claude-sonnet-4`"),
     make_live: boolean2().default(false).describe("When `true`, immediately point the evaluator's live version at this new version"),
-    output_config: OutputConfig$zodSchema.nullable().optional().describe("The scale points and their labels. Required for a `rating` evaluator. A `binary` evaluator uses the default Correct/Wrong labels unless you set your own"),
+    output_config: OutputConfig$zodSchema.nullable().optional().describe("The rows of the rating scale, each a number and its label. Required for a `rating` evaluator. For a rating evaluator, give the label for the worst result the lowest number and the label for the best result the highest number. A `binary` evaluator uses the default Correct/Wrong labels unless you set your own"),
     system_prompt: string2().describe('Judge system prompt. May contain `{{"{{"}}variable}}` placeholders'),
     variables: array(VariableSpec$zodSchema).nullable().optional().describe("Declared prompt variables. Omit if the prompt has none. After the first version the variable names are fixed. You can change a variable's description or default, but not add, remove, or rename one")
   }).describe("Request body for adding a version to an existing evaluator.");
@@ -61565,7 +61571,7 @@ var init_evaluatorversioncompact = __esm(() => {
   EvaluatorVersionCompact$zodSchema = object({
     created_at: string2().describe("When the version was created (ISO 8601 UTC)"),
     judge_model: string2().describe("The model that runs the judge, named the way its provider does, for example `openai/gpt-4.1` or `anthropic/claude-sonnet-4`"),
-    output_config: OutputConfig$zodSchema.nullable().optional().describe("The scale points and their labels. Required for a `rating` evaluator. A `binary` evaluator uses the default Correct/Wrong labels unless you set your own"),
+    output_config: OutputConfig$zodSchema.nullable().optional().describe("The rows of the rating scale, each a number and its label. Required for a `rating` evaluator. For a rating evaluator, give the label for the worst result the lowest number and the label for the best result the highest number. A `binary` evaluator uses the default Correct/Wrong labels unless you set your own"),
     system_prompt: string2().nullable().optional().describe('Judge system prompt, with `{{"{{"}}variable}}` placeholders unrendered'),
     uuid: string2().describe("Version ID"),
     variables: array(VariableSpec$zodSchema).nullable().optional().describe("Declared prompt variables"),
@@ -63387,7 +63393,7 @@ and metadata.
 function createMCPServer(deps) {
   const server = new McpServer({
     name: "CalibrateMcp",
-    version: "0.0.92"
+    version: "0.0.94"
   });
   const getClient = deps.getSDK || (() => new CalibrateMcpCore({
     security: deps.security,
@@ -64725,7 +64731,7 @@ http_headers = { "api-key-auth" = "YOUR_API_KEY_AUTH" }`;
         <h1>Instructions</h1>
         <p>One-click installation for Claude Desktop users</p>
         <div class="instruction-item">
-          <a href="https://github.com/dalmia/calibrate-mcp/releases/download/v0.0.92/mcp-server.mcpb" download="mcp-server.mcpb" class="action-button header-action" style="display: inline-flex; margin-bottom: 16px;">
+          <a href="https://github.com/dalmia/calibrate-mcp/releases/download/v0.0.94/mcp-server.mcpb" download="mcp-server.mcpb" class="action-button header-action" style="display: inline-flex; margin-bottom: 16px;">
             \uD83D\uDCE5 Download MCP Bundle
           </a>
         </div>
@@ -67679,7 +67685,7 @@ var routes = buildRouteMap({
 var app = buildApplication(routes, {
   name: "mcp",
   versionInfo: {
-    currentVersion: "0.0.92"
+    currentVersion: "0.0.94"
   }
 });
 run(app, process3.argv.slice(2), buildContext(process3));
@@ -67687,5 +67693,5 @@ export {
   app
 };
 
-//# debugId=C5E7357E5C5AD2E264756E2164756E21
+//# debugId=D36DA465B26C45F864756E2164756E21
 //# sourceMappingURL=mcp-server.js.map