@wix/pathgrade 0.29.0

package/package.json

@@ -0,0 +1,88 @@
+{
+  "name": "@wix/pathgrade",
+  "version": "0.29.0",
+  "description": "Evaluate whether AI agents discover and use your skills correctly",
+  "exports": {
+    ".": {
+      "types": "./dist/sdk/index.d.ts",
+      "default": "./dist/sdk/index.js"
+    },
+    "./plugin": {
+      "types": "./dist/plugin/index.d.ts",
+      "default": "./dist/plugin/index.js"
+    },
+    "./mcp-mock": {
+      "types": "./dist/core/mcp-mock.d.ts",
+      "default": "./dist/core/mcp-mock.js"
+    },
+    "./package.json": "./package.json"
+  },
+  "bin": "./bin/pathgrade.js",
+  "files": [
+    "dist/**/*.js",
+    "dist/**/*.d.ts",
+    "dist/viewer.html",
+    "!dist/**/*.js.map",
+    "!dist/**/*.d.ts.map",
+    "bin/",
+    "templates/",
+    "README.md",
+    "LICENSE"
+  ],
+  "scripts": {
+    "test": "tsc && vitest run",
+    "test:coverage": "vitest run --coverage",
+    "dev": "ts-node src/pathgrade.ts",
+    "build": "tsc -p tsconfig.build.json && cp src/viewer.html dist/viewer.html"
+  },
+  "keywords": [
+    "ai",
+    "agent",
+    "skills",
+    "evaluation",
+    "benchmark",
+    "gemini",
+    "claude",
+    "codex",
+    "openai",
+    "llm",
+    "testing"
+  ],
+  "publishConfig": {
+    "registry": "https://registry.npmjs.org/",
+    "access": "public"
+  },
+  "wix": {
+    "artifact": {
+      "groupId": "com.wixpress",
+      "artifactId": "pathgrade"
+    }
+  },
+  "author": "Nadav Lachish",
+  "license": "MIT",
+  "type": "module",
+  "engines": {
+    "node": ">=20.11.0"
+  },
+  "peerDependencies": {
+    "vitest": "^4.0.0"
+  },
+  "devDependencies": {
+    "@types/fs-extra": "^11.0.4",
+    "@types/picomatch": "^4.0.2",
+    "@vitest/coverage-v8": "^4.0.18",
+    "ts-node": "^10.9.2",
+    "tsx": "^4.21.0",
+    "vitest": "^4.0.18"
+  },
+  "dependencies": {
+    "@anthropic-ai/claude-agent-sdk": "0.2.85",
+    "@types/node": "^25.3.1",
+    "fs-extra": "^11.3.3",
+    "jiti": "^2.6.1",
+    "picomatch": "^4.0.4",
+    "typescript": "^5.9.3",
+    "zod": "4.3.6"
+  },
+  "falconPackageHash": "d5f15991255bd9d6a592b9cf3624be70aea80aedb73561fee73bef2e"
+}

package/templates/eval.ts.template

@@ -0,0 +1,28 @@
+import { describe, it, expect } from 'vitest';
+import { createAgent, check, judge, evaluate } from '@wix/pathgrade';
+
+describe('{{TASK_NAME}}', () => {
+  it('scores the agent result', async () => {
+    const agent = await createAgent({
+      timeout: 300,
+      skillDir: __dirname,
+      // workspace: path.join(__dirname, 'test', 'fixtures', 'my-fixture'),
+    });
+
+    try {
+      await agent.prompt(`{{INSTRUCTION}}`);
+
+      const result = await evaluate(agent, [
+        check('TODO: add deterministic scorer', async () => false),
+        judge('TODO: add rubric scorer', {
+          rubric: `TODO: Write evaluation criteria.`,
+          weight: 0.3,
+        }),
+      ]);
+
+      expect(result.score).toBeGreaterThanOrEqual(0);
+    } finally {
+      await agent.dispose();
+    }
+  });
+});