agentv 3.10.2 → 3.10.3

package/dist/{chunk-6UE665XI.js → chunk-7LC3VNOC.js}

@@ -17,7 +17,7 @@ import {
   validateEvalFile,
   validateFileReferences,
   validateTargetsFile
-} from "./chunk-F7LAJMTO.js";
+} from "./chunk-U556GRI3.js";
 import {
   createBuiltinRegistry,
   createProvider,
@@ -35,7 +35,7 @@ import {
   toSnakeCaseDeep as toSnakeCaseDeep2,
   transpileEvalYamlFile,
   trimBaselineResult
-} from "./chunk-KGK5NUFG.js";
+} from "./chunk-JUQCB3ZW.js";
 import {
   __commonJS,
   __esm,
@@ -4187,7 +4187,7 @@ var evalRunCommand = command({
   },
   handler: async (args) => {
     if (args.evalPaths.length === 0 && process.stdin.isTTY) {
-      const { launchInteractiveWizard } = await import("./interactive-EO6AR2R3.js");
+      const { launchInteractiveWizard } = await import("./interactive-CSA4KIND.js");
       await launchInteractiveWizard();
       return;
     }
@@ -6287,4 +6287,4 @@ export {
   preprocessArgv,
   runCli
 };
-//# sourceMappingURL=chunk-6UE665XI.js.map
+//# sourceMappingURL=chunk-7LC3VNOC.js.map

package/dist/{chunk-KGK5NUFG.js → chunk-JUQCB3ZW.js}

@@ -19552,7 +19552,7 @@ ${basePrompt}` : basePrompt;
     if (itemType === "command_execution") {
       completedToolCalls.push({
         tool: "command_execution",
-        input: item.command,
+        input: { command: item.command },
         output: item.aggregated_output,
         id: item.id
       });
@@ -20383,11 +20383,22 @@ async function loadCopilotSdk() {
     try {
       copilotSdkModule = await import("@github/copilot-sdk");
     } catch (error) {
+      const message = error instanceof Error ? error.message : String(error);
+      if (message.includes("vscode-jsonrpc")) {
+        throw new Error(
+          `Failed to load @github/copilot-sdk due to a known ESM compatibility issue with vscode-jsonrpc (https://github.com/github/copilot-sdk/issues/710).
+
+Workarounds:
+  - Use the copilot-cli target instead (recommended): set target type to "copilot-cli" in your eval YAML
+  - If running under Node.js 24+: set NODE_OPTIONS="--experimental-specifier-resolution=node"
+  - Wait for vscode-jsonrpc@9.0.0 stable to be released upstream`
+        );
+      }
       throw new Error(
         `Failed to load @github/copilot-sdk. Please install it:
   npm install @github/copilot-sdk
 
-Original error: ${error instanceof Error ? error.message : String(error)}`
+Original error: ${message}`
       );
     }
   }
@@ -21592,6 +21603,13 @@ function extractToolCalls4(content) {
         id: typeof p.id === "string" ? p.id : void 0
       });
     }
+    if (p.type === "toolCall" && typeof p.name === "string") {
+      toolCalls.push({
+        tool: p.name,
+        input: p.arguments,
+        id: typeof p.id === "string" ? p.id : void 0
+      });
+    }
     if (p.type === "tool_result" && typeof p.tool_use_id === "string") {
       const existing = toolCalls.find((tc) => tc.id === p.tool_use_id);
       if (existing) {
@@ -25778,12 +25796,31 @@ var COPILOT_MATCHER = {
   readToolPrefixes: ["Viewing "],
   readInputFields: ["file_path", "path"]
 };
+var PI_CODING_AGENT_MATCHER = {
+  skillTools: [],
+  skillInputField: "skill",
+  readTools: ["read"],
+  readInputField: "path",
+  readInputFields: ["path", "file_path", "filePath"]
+};
+var CODEX_MATCHER = {
+  skillTools: [],
+  skillInputField: "skill",
+  readTools: ["command_execution"],
+  readInputField: "command",
+  skillToolPrefixes: ["mcp:"],
+  readToolPrefixes: ["mcp:"],
+  readInputFields: ["command", "path", "file_path", "filePath"]
+};
 var PROVIDER_TOOL_SEMANTICS = {
   claude: CLAUDE_MATCHER,
   "claude-cli": CLAUDE_MATCHER,
   "claude-sdk": CLAUDE_MATCHER,
-  "pi-coding-agent": CLAUDE_MATCHER,
-  "pi-agent-sdk": CLAUDE_MATCHER,
+  codex: CODEX_MATCHER,
+  "pi-coding-agent": PI_CODING_AGENT_MATCHER,
+  // pi-agent-sdk has no tools, so skill detection is a no-op. Kept for completeness.
+  // TODO: consider removing pi-agent-sdk provider entirely.
+  "pi-agent-sdk": PI_CODING_AGENT_MATCHER,
   "copilot-cli": COPILOT_MATCHER,
   "copilot-sdk": COPILOT_MATCHER,
   vscode: COPILOT_MATCHER,
@@ -25807,33 +25844,37 @@ var SkillTriggerEvaluator = class {
     const shouldTrigger = this.config.should_trigger !== false;
     const providerKind = context2.provider?.kind;
     const matcher = this.resolveMatcher(providerKind);
-    const firstTool = (context2.output ?? []).flatMap((msg) => msg.toolCalls ?? [])[0];
+    const allToolCalls = (context2.output ?? []).flatMap((msg) => msg.toolCalls ?? []);
     let triggered = false;
     let evidence = "";
-    if (firstTool) {
-      const input = firstTool.input ?? {};
-      if (matcher.skillTools.includes(firstTool.tool)) {
+    for (const toolCall of allToolCalls) {
+      const input = toolCall.input ?? {};
+      if (matcher.skillTools.includes(toolCall.tool)) {
         const skillArg = String(input[matcher.skillInputField] ?? "");
         if (skillArg.includes(skillName)) {
           triggered = true;
           evidence = `Skill tool invoked with ${matcher.skillInputField}="${skillArg}"`;
+          break;
         }
       } else if (matcher.skillToolPrefixes?.some(
-        (prefix) => firstTool.tool.startsWith(prefix) && firstTool.tool.includes(skillName)
+        (prefix) => toolCall.tool.startsWith(prefix) && toolCall.tool.includes(skillName)
       )) {
         triggered = true;
-        evidence = `Skill tool invoked via tool name "${firstTool.tool}"`;
-      } else if (matcher.readTools.includes(firstTool.tool)) {
+        evidence = `Skill tool invoked via tool name "${toolCall.tool}"`;
+        break;
+      } else if (matcher.readTools.includes(toolCall.tool)) {
         const filePath = this.readPathFromInput(input, matcher);
         if (filePath.includes(skillName)) {
           triggered = true;
           evidence = `Read tool loaded skill file: ${filePath}`;
+          break;
         }
       } else if (matcher.readToolPrefixes?.some(
-        (prefix) => firstTool.tool.startsWith(prefix) && firstTool.tool.includes(skillName)
+        (prefix) => toolCall.tool.startsWith(prefix) && toolCall.tool.includes(skillName)
       )) {
         triggered = true;
-        evidence = `Read tool loaded skill file via tool name "${firstTool.tool}"`;
+        evidence = `Read tool loaded skill file via tool name "${toolCall.tool}"`;
+        break;
       }
     }
     const pass = triggered === shouldTrigger;
@@ -25855,7 +25896,7 @@ var SkillTriggerEvaluator = class {
       verdict: "fail",
       assertions: [
         {
-          text: shouldTrigger ? firstTool ? `First tool was "${firstTool.tool}" \u2014 not a skill/read tool for "${skillName}"` : "No tool calls recorded" : evidence || `Skill "${skillName}" triggered unexpectedly`,
+          text: shouldTrigger ? allToolCalls.length > 0 ? `Skill "${skillName}" not found in ${allToolCalls.length} tool call(s)` : "No tool calls recorded" : evidence || `Skill "${skillName}" triggered unexpectedly`,
           passed: false
         }
       ],
@@ -31062,4 +31103,4 @@ export {
   OtelStreamingObserver,
   createAgentKernel
 };
-//# sourceMappingURL=chunk-KGK5NUFG.js.map
+//# sourceMappingURL=chunk-JUQCB3ZW.js.map