@dalmia/calibrate-mcp 0.0.17 → 0.0.18

package/bin/mcp-server.js

@@ -51392,9 +51392,9 @@ var init_config = __esm(() => {
   SDK_METADATA = {
     language: "typescript",
     openapiDocVersion: "0.1.0",
-    sdkVersion: "0.0.17",
+    sdkVersion: "0.0.18",
     genVersion: "2.915.1",
-    userAgent: "speakeasy-sdk/mcp-typescript 0.0.17 2.915.1 0.1.0 @dalmia/calibrate-mcp"
+    userAgent: "speakeasy-sdk/mcp-typescript 0.0.18 2.915.1 0.1.0 @dalmia/calibrate-mcp"
   };
 });
 
@@ -54672,14 +54672,41 @@ var init_agentTestsListForAgent2 = __esm(() => {
   };
 });
 
+// src/models/modelrunsummary.ts
+var ModelRunSummary$zodSchema;
+var init_modelrunsummary = __esm(() => {
+  init_zod();
+  ModelRunSummary$zodSchema = object({
+    failed: int().nullable().optional().describe("Number of test cases that failed for this model"),
+    message: string2().default("").describe("Status or result message for this model"),
+    model: string2().describe("Model name these results are for"),
+    passed: int().nullable().optional().describe("Number of test cases that passed for this model"),
+    success: boolean2().nullable().optional().describe("Whether this model's run succeeded"),
+    total_tests: int().nullable().optional().describe("Total test cases for this model")
+  }).describe("Flat summary for one model in a benchmark run-LIST item. The full results\nfor each case of a model live on the benchmark detail endpoint\n(`GET /agent-tests/benchmark/{task_id}`), not here.");
+});
+
+// src/models/testruncasesummary.ts
+var TestRunCaseSummary$zodSchema;
+var init_testruncasesummary = __esm(() => {
+  init_zod();
+  TestRunCaseSummary$zodSchema = object({
+    name: string2().nullable().optional().describe("Name of the test case"),
+    passed: boolean2().nullable().optional().describe("Whether the case passed (null if it errored or is still running)")
+  }).describe(`Flat summary for one test case in the run-LIST endpoints. Carries only
+enough to render a run's pass/fail breakdown and a case name. The full detail
+for each case (agent output, judge verdicts, reasoning, latency, cost, the
+test-case definition) lives on the run-DETAIL endpoint
+(\`GET /agent-tests/run/{task_id}\`).`);
+});
+
 // src/models/agenttestrunlistitem.ts
 var AgentTestRunListItemType$zodSchema, AgentTestRunListItem$zodSchema;
 var init_agenttestrunlistitem = __esm(() => {
   init_zod();
-  init_modelresult();
+  init_modelrunsummary();
   init_taskstatus();
-  init_testcaseresult();
-  init_testrunevaluator();
+  init_testruncasesummary();
   AgentTestRunListItemType$zodSchema = _enum([
     "llm-unit-test",
     "llm-benchmark"
@@ -54687,15 +54714,13 @@ var init_agenttestrunlistitem = __esm(() => {
   AgentTestRunListItem$zodSchema = object({
     cost: record(string2(), any()).nullable().optional().describe("Aggregated cost as `{mean, min, max, count}` (USD)"),
     error: boolean2().default(false).describe("True if the run failed"),
-    evaluators: array(TestRunEvaluator$zodSchema).nullable().optional().describe("The evaluators used in this run. Each verdict in `judge_results` links to one of these by `evaluator_uuid`"),
     failed: int().nullable().optional().describe("Number of test cases that failed"),
     is_public: boolean2().default(false).describe("Whether the run is shared publicly"),
     latency_ms: record(string2(), any()).nullable().optional().describe("Aggregated latency in milliseconds, as `{p50, p95, p99, count}`"),
-    leaderboard_summary: array(record(string2(), any())).nullable().optional().describe("Leaderboard comparing the models, one row per model. Columns vary by benchmark: a `model` column plus pass/fail counts, latency, cost, and one score column per evaluator, keyed by evaluator name"),
-    model_results: array(ModelResult$zodSchema).nullable().optional().describe("Results for each model in a benchmark run"),
+    model_results: array(ModelRunSummary$zodSchema).nullable().optional().describe("Flat summary for each model in a benchmark run (fetch the benchmark detail for full results)"),
     name: string2().describe("Display name, such as `Run 1` for a unit test or `Benchmark 1` for a benchmark"),
     passed: int().nullable().optional().describe("Number of test cases that passed"),
-    results: array(TestCaseResult$zodSchema).nullable().optional().describe("Results for each test case"),
+    results: array(TestRunCaseSummary$zodSchema).nullable().optional().describe("Flat pass/fail summary for each test case (fetch the run detail for full results)"),
     share_token: string2().nullable().optional().describe("Token for building the public share URL"),
     status: TaskStatus$zodSchema,
     total_tests: int().nullable().optional().describe("Total number of test cases"),
@@ -58456,7 +58481,7 @@ fields to change (name, config). Returns the updated test. Fails with
 function createMCPServer(deps) {
   const server = new McpServer({
     name: "CalibrateMcp",
-    version: "0.0.17"
+    version: "0.0.18"
   });
   const getClient = deps.getSDK || (() => new CalibrateMcpCore({
     security: deps.security,
@@ -59662,7 +59687,7 @@ 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/v0.0.17/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/v0.0.18/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>
@@ -62314,7 +62339,7 @@ var routes = buildRouteMap({
 var app = buildApplication(routes, {
   name: "mcp",
   versionInfo: {
-    currentVersion: "0.0.17"
+    currentVersion: "0.0.18"
   }
 });
 run(app, process4.argv.slice(2), buildContext(process4));
@@ -62322,5 +62347,5 @@ export {
   app
 };
 
-//# debugId=312531E882E78CEA64756E2164756E21
+//# debugId=E95251CDECEB21E364756E2164756E21
 //# sourceMappingURL=mcp-server.js.map