agentv 2.12.0 → 2.14.0-next.1

package/dist/{chunk-6KU2ZUFJ.js → chunk-ZSSGXZX6.js}

@@ -2,6 +2,7 @@ import { createRequire } from 'node:module'; const require = createRequire(impor
 import {
   detectFileType,
   findRepoRoot,
+  package_default,
   resolveEvalPaths,
   runEvalCommand,
   selectTarget,
@@ -10,7 +11,7 @@ import {
   validateEvalFile,
   validateFileReferences,
   validateTargetsFile
-} from "./chunk-YBJX5CP6.js";
+} from "./chunk-K2APOWTE.js";
 import {
   RepoManager,
   assembleLlmJudgePrompt,
@@ -25,7 +26,7 @@ import {
   toCamelCaseDeep,
   toSnakeCaseDeep,
   trimBaselineResult
-} from "./chunk-LUHCYBMD.js";
+} from "./chunk-OQN2GDEU.js";
 import {
   __commonJS,
   __esm,
@@ -2875,56 +2876,6 @@ function oneOf(literals) {
   };
 }
 
-// package.json
-var package_default = {
-  name: "agentv",
-  version: "2.12.0",
-  description: "CLI entry point for AgentV",
-  type: "module",
-  repository: {
-    type: "git",
-    url: "https://github.com/EntityProcess/agentv.git"
-  },
-  homepage: "https://github.com/EntityProcess/agentv#readme",
-  bugs: {
-    url: "https://github.com/EntityProcess/agentv/issues"
-  },
-  bin: {
-    agentv: "./dist/cli.js"
-  },
-  files: ["dist", "README.md"],
-  scripts: {
-    dev: "bun src/cli.ts",
-    build: "tsup && bun run copy-readme",
-    "copy-readme": `bun -e "import { cpSync } from 'fs'; cpSync('../../README.md', 'README.md')"`,
-    prepublishOnly: "bun run copy-readme",
-    typecheck: "tsc --noEmit",
-    lint: "biome check .",
-    format: "biome format --write .",
-    fix: "biome check --write .",
-    test: "bun test",
-    "test:watch": "bun test --watch"
-  },
-  dependencies: {
-    "@anthropic-ai/claude-agent-sdk": "^0.2.49",
-    "@github/copilot-sdk": "^0.1.25",
-    "@inquirer/prompts": "^8.2.1",
-    "@mariozechner/pi-agent-core": "^0.54.2",
-    "@mariozechner/pi-ai": "^0.54.2",
-    "@openai/codex-sdk": "^0.104.0",
-    "cmd-ts": "^0.14.3",
-    dotenv: "^16.4.5",
-    "fast-glob": "^3.3.3",
-    json5: "^2.2.3",
-    micromatch: "^4.0.8",
-    yaml: "^2.6.1"
-  },
-  devDependencies: {
-    "@agentv/core": "workspace:*",
-    execa: "^9.3.0"
-  }
-};
-
 // src/commands/cache/add.ts
 import { existsSync } from "node:fs";
 import { join, resolve } from "node:path";
@@ -3533,7 +3484,7 @@ tests:
     input: "Hello, how are you?"
     expected_output: "I'm doing well, thank you for asking!"
     assert:
-      - type: llm_judge
+      - type: llm-judge
         rubric:
           accuracy:
             weight: 0.6
@@ -3812,7 +3763,7 @@ var evalPromptJudgeCommand = command({
       });
       outputs.push({
         name: "default_llm_judge",
-        type: "llm_judge",
+        type: "llm-judge",
         status: "prompt_ready",
         prompt: {
           system_prompt: assembly.systemPrompt,
@@ -3830,7 +3781,7 @@ var evalPromptJudgeCommand = command({
 });
 async function processEvaluator(config, evalCase, candidate, promptInputs) {
   switch (config.type) {
-    case "code": {
+    case "code-judge": {
       const codeConfig = config;
       const script = codeConfig.command ?? codeConfig.script ?? [];
       const scriptCwd = codeConfig.resolvedCwd ?? codeConfig.cwd;
@@ -3855,14 +3806,14 @@ async function processEvaluator(config, evalCase, candidate, promptInputs) {
         const parsed = JSON.parse(stdout);
         return {
           name: codeConfig.name,
-          type: "code_judge",
+          type: "code-judge",
           status: "completed",
           result: parsed
         };
       } catch (error) {
         return {
           name: codeConfig.name,
-          type: "code_judge",
+          type: "code-judge",
           status: "completed",
           result: {
             score: 0,
@@ -3871,7 +3822,7 @@ async function processEvaluator(config, evalCase, candidate, promptInputs) {
         };
       }
     }
-    case "llm_judge": {
+    case "llm-judge": {
       const llmConfig = config;
       const assembly = assembleLlmJudgePrompt({
         evalCase,
@@ -3881,7 +3832,7 @@ async function processEvaluator(config, evalCase, candidate, promptInputs) {
       });
       return {
         name: llmConfig.name,
-        type: "llm_judge",
+        type: "llm-judge",
         status: "prompt_ready",
         prompt: {
           system_prompt: assembly.systemPrompt,
@@ -3947,7 +3898,7 @@ var evalPromptOverviewCommand = command({
       "",
       "The output contains an `evaluators` array. Each evaluator has a `status`:",
       "",
-      '- **`"completed"`** \u2014 Score is final (code_judge ran deterministically). Read `result.score` (0.0\u20131.0).',
+      '- **`"completed"`** \u2014 Score is final (code-judge ran deterministically). Read `result.score` (0.0\u20131.0).',
       '- **`"prompt_ready"`** \u2014 LLM grading required. Send `prompt.system_prompt` as system and',
       "  `prompt.user_prompt` as user to your LLM. Parse the JSON response to get `score`, `hits`, `misses`.",
       ""
@@ -4123,11 +4074,20 @@ var evalRunCommand = command({
     otelGroupTurns: flag({
       long: "otel-group-turns",
       description: "Group messages into turn spans for multi-turn evaluations (requires --export-otel)"
+    }),
+    retryErrors: option({
+      type: optional(string),
+      long: "retry-errors",
+      description: "Path to previous output JSONL \u2014 re-run only execution_error test cases"
+    }),
+    strict: flag({
+      long: "strict",
+      description: "Exit with error on version mismatch (instead of warning)"
     })
   },
   handler: async (args) => {
     if (args.evalPaths.length === 0 && process.stdin.isTTY) {
-      const { launchInteractiveWizard } = await import("./interactive-TOUKPSHP.js");
+      const { launchInteractiveWizard } = await import("./interactive-WF6UO63B.js");
       await launchInteractiveWizard();
       return;
     }
@@ -4157,7 +4117,9 @@ var evalRunCommand = command({
       exportOtel: args.exportOtel,
       otelBackend: args.otelBackend,
       otelCaptureContent: args.otelCaptureContent,
-      otelGroupTurns: args.otelGroupTurns
+      otelGroupTurns: args.otelGroupTurns,
+      retryErrors: args.retryErrors,
+      strict: args.strict
     };
     await runEvalCommand({ testFiles: resolvedPaths, rawOptions });
   }
@@ -4758,12 +4720,12 @@ var traceListCommand = command({
 var SUPPORTED_TYPES = [
   "contains",
   "regex",
-  "is_json",
+  "is-json",
   "equals",
   "latency",
   "cost",
-  "token_usage",
-  "execution_metrics"
+  "token-usage",
+  "execution-metrics"
 ];
 function parseKeyValues(s) {
   const result = {};
@@ -4777,7 +4739,7 @@ function parseKeyValues(s) {
 }
 function parseAssertSpec(spec) {
   const colonIdx = spec.indexOf(":");
-  const type = colonIdx === -1 ? spec : spec.slice(0, colonIdx);
+  const type = (colonIdx === -1 ? spec : spec.slice(0, colonIdx)).replace(/_/g, "-");
   const params = colonIdx === -1 ? "" : spec.slice(colonIdx + 1);
   switch (type) {
     case "contains":
@@ -4786,8 +4748,8 @@ function parseAssertSpec(spec) {
     case "regex":
       if (!params) throw new Error("regex requires a pattern: regex:<pattern>");
       return { name: "regex", type: "regex", value: params };
-    case "is_json":
-      return { name: "is_json", type: "is_json" };
+    case "is-json":
+      return { name: "is-json", type: "is-json" };
     case "equals":
       if (!params) throw new Error("equals requires a value: equals:<value>");
       return { name: "equals", type: "equals", value: params };
@@ -4803,19 +4765,19 @@ function parseAssertSpec(spec) {
         throw new Error("cost requires a budget in USD: cost:<usd>");
       return { name: "cost", type: "cost", budget };
     }
-    case "token_usage": {
+    case "token-usage": {
       const kv = parseKeyValues(params);
-      const config = { name: "token_usage", type: "token_usage" };
+      const config = { name: "token-usage", type: "token-usage" };
       if (kv.max_total) config.max_total = Number(kv.max_total);
       if (kv.max_input) config.max_input = Number(kv.max_input);
       if (kv.max_output) config.max_output = Number(kv.max_output);
       return config;
     }
-    case "execution_metrics": {
+    case "execution-metrics": {
       const kv = parseKeyValues(params);
       const config = {
-        name: "execution_metrics",
-        type: "execution_metrics"
+        name: "execution-metrics",
+        type: "execution-metrics"
       };
       if (kv.max_tool_calls) config.max_tool_calls = Number(kv.max_tool_calls);
       if (kv.max_llm_calls) config.max_llm_calls = Number(kv.max_llm_calls);
@@ -4861,7 +4823,7 @@ var stubProvider = {
   }
 };
 var stubLlmJudge = {
-  kind: "llm_judge",
+  kind: "llm-judge",
   evaluate() {
     throw new Error("trace score does not support LLM-based evaluators");
   }
@@ -4955,7 +4917,7 @@ var traceScoreCommand = command({
       type: string,
       long: "assert",
       short: "a",
-      description: "Evaluator spec: contains:<val>, regex:<pat>, is_json, equals:<val>, latency:<ms>, cost:<usd>, token_usage:<params>, execution_metrics:<params>"
+      description: "Evaluator spec: contains:<val>, regex:<pat>, is-json, equals:<val>, latency:<ms>, cost:<usd>, token-usage:<params>, execution-metrics:<params>"
     }),
     testId: option({
       type: optional(string),
@@ -4990,7 +4952,7 @@ var traceScoreCommand = command({
       console.error(`${c2.yellow}Warning:${c2.reset} No results found in ${file}`);
       process.exit(0);
     }
-    const traceRequired = ["latency", "cost", "token_usage", "execution_metrics"].includes(
+    const traceRequired = ["latency", "cost", "token-usage", "execution-metrics"].includes(
       evaluatorConfig.type
     );
     if (traceRequired) {
@@ -5831,4 +5793,4 @@ export {
   preprocessArgv,
   runCli
 };
-//# sourceMappingURL=chunk-6KU2ZUFJ.js.map
+//# sourceMappingURL=chunk-ZSSGXZX6.js.map