@gleanwork/mcp-server-tester 1.0.0-beta.3 → 1.0.0-beta.4

package/README.md

@@ -1,6 +1,6 @@
 # @gleanwork/mcp-server-tester
 
-[![Experimental](https://img.shields.io/badge/-Experimental-D8FD49?style=flat-square&logo=data:image/svg+xml;base64,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&labelColor=343CED)](https://github.com/gleanwork/.github/blob/main/docs/repository-stability.md#experimental)
+[![GA](https://img.shields.io/badge/-GA-F6F3EB?style=flat-square&logo=data:image/svg+xml;base64,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&labelColor=343CED)](https://github.com/gleanwork/.github/blob/main/docs/repository-stability.md#ga)
 [![npm version](https://img.shields.io/npm/v/@gleanwork/mcp-server-tester)](https://www.npmjs.com/package/@gleanwork/mcp-server-tester)
 [![CI](https://github.com/gleanwork/mcp-server-tester/actions/workflows/ci.yml/badge.svg)](https://github.com/gleanwork/mcp-server-tester/actions/workflows/ci.yml)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
@@ -11,7 +11,7 @@ A testing and evaluation framework for [Model Context Protocol (MCP)](https://mo
 
 The `mcp` Playwright fixture connects to your MCP server (stdio or HTTP) and exposes a high-level API for calling tools and asserting responses. Custom matchers keep assertions readable.
 
-```typescript
+```typescript snippet=snippets/basic-test.ts
 import { test, expect } from '@gleanwork/mcp-server-tester/fixtures/mcp';
 
 test('read_file returns file contents', async ({ mcp }) => {
@@ -48,7 +48,7 @@ Available matchers:
 
 Eval datasets let you define test cases as JSON files and run them with `runEvalDataset()`. Each case specifies a tool call and one or more assertions.
 
-```json
+```json snippet=snippets/eval-dataset.json
 {
   "name": "file-ops",
   "cases": [
@@ -73,7 +73,7 @@ Eval datasets let you define test cases as JSON files and run them with `runEval
 }
 ```
 
-```typescript
+```typescript snippet=snippets/run-eval-dataset.ts
 import { test, expect } from '@gleanwork/mcp-server-tester/fixtures/mcp';
 import { loadEvalDataset, runEvalDataset } from '@gleanwork/mcp-server-tester';
 import { z } from 'zod';
@@ -102,12 +102,12 @@ Supported assertion types:
 
 In LLM host mode, a real LLM receives your server's tool list and a natural language prompt, then decides which tools to call. This tests whether your tool names, descriptions, and input schemas are clear enough for autonomous use — a different question from whether the tools return correct output.
 
-```json
+```json snippet=snippets/mcp-host-dataset.json
 {
   "id": "find-config",
-  "mode": "llm_host",
+  "mode": "mcp_host",
   "scenario": "Find the application config file and return its contents",
-  "llmHostConfig": {
+  "mcpHostConfig": {
     "provider": "anthropic",
     "model": "claude-opus-4-20250514"
   },
@@ -119,7 +119,7 @@ In LLM host mode, a real LLM receives your server's tool list and a natural lang
 }
 ```
 
-LLM host mode makes real API calls and produces non-deterministic results. Use `iterations` to run a case multiple times and measure pass rate rather than expecting 100% on a single run. See the [LLM Host Guide](docs/llm-host.md) for configuration and cost management.
+LLM host mode makes real API calls and produces non-deterministic results. Use `iterations` to run a case multiple times and measure pass rate rather than expecting 100% on a single run. See the [LLM Host Guide](docs/mcp-host.md) for configuration and cost management.
 
 ## Installation
 
@@ -147,7 +147,7 @@ The CLI wizard creates a `playwright.config.ts`, example tests, and a sample eva
 
 Point the framework at your MCP server in `playwright.config.ts`:
 
-```typescript
+```typescript snippet=snippets/playwright-config.ts
 import { defineConfig } from '@playwright/test';
 
 export default defineConfig({
@@ -174,12 +174,13 @@ For HTTP servers, set `transport: 'http'` and `serverUrl`. For servers that requ
 
 - [Quick Start](./docs/quickstart.md) — detailed setup and configuration
 - [Expectations](./docs/expectations.md) — all assertion types including snapshot sanitizers
-- [LLM Host Simulation](docs/llm-host.md) — tool discoverability testing
+- [LLM Host Simulation](docs/mcp-host.md) — tool discoverability testing
 - [API Reference](./docs/api-reference.md)
 - [Transports](./docs/transports.md) — stdio and HTTP configuration, OAuth
 - [CLI Commands](./docs/cli.md) — init, generate, login, token
 - [UI Reporter](./docs/ui-reporter.md) — interactive web UI for test results
 - [Development](./docs/development.md) — contributing and building
+- [Migration Guide (v0.12 → v1.0)](./docs/migrations/migration-1.0.md) — upgrading from pre-1.0 releases
 
 ## Examples
 

package/dist/cli/index.js

@@ -17,7 +17,7 @@ import { SSEClientTransport } from '@modelcontextprotocol/sdk/client/sse.js';
 import { z } from 'zod';
 import createDebug from 'debug';
 import { ProxyAgent, Agent } from 'undici';
-import { readFileSync } from 'fs';
+import { existsSync, readFileSync } from 'fs';
 import * as oauth from 'oauth4webapi';
 import { homedir } from 'os';
 import * as http from 'http';
@@ -80,7 +80,7 @@ function JsonPreview({ data, maxLines = 15 }) {
 
 // package.json
 var package_default = {
-  version: "1.0.0-beta.3"};
+  version: "1.0.0-beta.4"};
 
 // src/cli/templates/index.ts
 function getPlaywrightConfigTemplate(answers) {
@@ -255,10 +255,10 @@ function getPackageJsonTemplate(projectName) {
     "evals"
   ],
   "dependencies": {
-    "@modelcontextprotocol/sdk": "^1.0.4",
+    "@modelcontextprotocol/sdk": "^1.27.0",
     "@playwright/test": "^1.49.0",
     "@gleanwork/mcp-server-tester": "^${package_default.version}",
-    "zod": "^3.24.1"
+    "zod": "^4.0.0"
   },
   "devDependencies": {
     "typescript": "^5.7.2"
@@ -531,7 +531,7 @@ async function init(options) {
   await waitUntilExit();
 }
 var MCPHostCapabilitiesSchema = z.object({
-  sampling: z.record(z.unknown()).optional(),
+  sampling: z.record(z.string(), z.unknown()).optional(),
   roots: z.object({
     listChanged: z.boolean()
   }).optional()
@@ -590,7 +590,7 @@ var HttpConfigSchema = z.object({
     }
     return true;
   }),
-  headers: z.record(z.string()).optional(),
+  headers: z.record(z.string(), z.string()).optional(),
   capabilities: MCPHostCapabilitiesSchema.optional(),
   connectTimeoutMs: z.number().positive().optional(),
   requestTimeoutMs: z.number().positive().optional(),
@@ -827,7 +827,7 @@ async function retryWithBackoff(fn, maxAttempts) {
           delayMs,
           err.message
         );
-        await new Promise((resolve3) => setTimeout(resolve3, delayMs));
+        await new Promise((resolve4) => setTimeout(resolve4, delayMs));
       } else {
         throw err;
       }
@@ -1683,7 +1683,7 @@ ${errorText}`
    */
   async startCallbackServer(expectedState) {
     const timeoutMs = this.config.timeoutMs ?? DEFAULT_TIMEOUT_MS;
-    return new Promise((resolve3, reject) => {
+    return new Promise((resolve4, reject) => {
       const server = http.createServer();
       const connections = /* @__PURE__ */ new Set();
       server.on("connection", (socket) => {
@@ -1756,7 +1756,7 @@ ${errorText}`
       server.listen(preferredPort, "127.0.0.1", () => {
         const address = server.address();
         debug2("Callback server listening on port", address.port);
-        resolve3({ port: address.port, codePromise, close: forceClose });
+        resolve4({ port: address.port, codePromise, close: forceClose });
       });
       server.on("error", (err) => {
         reject(err);
@@ -1777,8 +1777,8 @@ ${errorText}`
       return;
     }
     try {
-      const open = await import('open');
-      await open.default(url.toString());
+      const open2 = await import('open');
+      await open2.default(url.toString());
       debug2("Opened browser for authentication");
     } catch (error) {
       debug2("Failed to open browser:", error);
@@ -3062,6 +3062,24 @@ async function token(serverUrl, options) {
   );
   await waitUntilExit();
 }
+async function open(options) {
+  const outputDir = resolve(options.dir ?? ".mcp-test-results");
+  const reportPath = join(outputDir, "latest", "index.html");
+  if (!existsSync(reportPath)) {
+    console.error(`No report found at ${reportPath}`);
+    console.error("Run your Playwright tests first to generate a report.");
+    process.exit(1);
+  }
+  console.log(`Opening report: ${reportPath}`);
+  try {
+    const { default: openBrowser } = await import('open');
+    await openBrowser(reportPath);
+  } catch (error) {
+    console.error("Failed to open report in browser:", error);
+    console.error(`Open manually: file://${reportPath}`);
+    process.exit(1);
+  }
+}
 
 // src/cli/index.ts
 var program = new Command();
@@ -3077,4 +3095,9 @@ program.command("token").description("Output stored OAuth tokens for CI/CD use")
   "Output format: env, json, or gh (default: env)",
   "env"
 ).option("--state-dir <dir>", "Custom directory for token storage").action(token);
+program.command("open").description("Open the MCP eval reporter UI in your browser").option(
+  "-d, --dir <directory>",
+  "Report output directory",
+  ".mcp-test-results"
+).action(open);
 program.parse();

package/dist/fixtures/mcp.d.ts

@@ -217,9 +217,9 @@ type RubricSpec = BuiltInRubric | {
 type ProviderKind = 'anthropic' | 'openai' | 'google';
 
 /**
- * Tool call validators for llm_host simulation results.
+ * Tool call validators for mcp_host simulation results.
  *
- * These validators extract the tool call trace from an LLMHostSimulationResult
+ * These validators extract the tool call trace from an MCPHostSimulationResult
  * and apply assertions against expected call lists and counts.
  */
 
@@ -408,7 +408,7 @@ declare global {
              */
             toSatisfyToolPredicate(predicate: ToolPredicate, description?: string): Promise<R>;
             /**
-             * Validates which tools the LLM called during an llm_host simulation.
+             * Validates which tools the LLM called during a mcp_host simulation.
              *
              * @example
              * ```typescript
@@ -420,7 +420,7 @@ declare global {
              */
             toHaveToolCalls(expectation: ToolCallExpectation): R;
             /**
-             * Validates the number of tool calls made during an llm_host simulation.
+             * Validates the number of tool calls made during a mcp_host simulation.
              *
              * @example
              * ```typescript
@@ -531,7 +531,7 @@ declare function toSatisfyToolPredicate(this: {
 /**
  * toHaveToolCalls Matcher
  *
- * Validates which tools the LLM called during an llm_host simulation.
+ * Validates which tools the LLM called during a mcp_host simulation.
  */
 
 /**
@@ -547,7 +547,7 @@ declare function toHaveToolCalls(this: {
 /**
  * toHaveToolCallCount Matcher
  *
- * Validates the number of tool calls made during an llm_host simulation.
+ * Validates the number of tool calls made during a mcp_host simulation.
  */
 
 /**

package/dist/fixtures/mcp.js

@@ -1215,7 +1215,7 @@ function validateToolCalls(response, expectation) {
   if (!isSimulationResult(response)) {
     return {
       pass: false,
-      message: "toolsTriggered expectation requires llm_host mode \u2014 response must be an LLMHostSimulationResult"
+      message: "toolsTriggered expectation requires mcp_host mode \u2014 response must be an MCPHostSimulationResult"
     };
   }
   const actual = response.toolCalls;
@@ -1275,7 +1275,7 @@ function validateToolCallCount(response, options) {
   if (!isSimulationResult(response)) {
     return {
       pass: false,
-      message: "toolCallCount expectation requires llm_host mode \u2014 response must be an LLMHostSimulationResult"
+      message: "toolCallCount expectation requires mcp_host mode \u2014 response must be an MCPHostSimulationResult"
     };
   }
   const count = response.toolCalls.length;
@@ -1337,7 +1337,7 @@ var expect = expect$1.extend({
   toHaveToolCallCount
 });
 var MCPHostCapabilitiesSchema = z.object({
-  sampling: z.record(z.unknown()).optional(),
+  sampling: z.record(z.string(), z.unknown()).optional(),
   roots: z.object({
     listChanged: z.boolean()
   }).optional()
@@ -1396,7 +1396,7 @@ var HttpConfigSchema = z.object({
     }
     return true;
   }),
-  headers: z.record(z.string()).optional(),
+  headers: z.record(z.string(), z.string()).optional(),
   capabilities: MCPHostCapabilitiesSchema.optional(),
   connectTimeoutMs: z.number().positive().optional(),
   requestTimeoutMs: z.number().positive().optional(),
@@ -1434,7 +1434,7 @@ var debugHttp = createDebug(`${NAMESPACE}:http`);
 
 // package.json
 var package_default = {
-  version: "1.0.0-beta.3"};
+  version: "1.0.0-beta.4"};
 var debug = createDebug("mcp-server-tester:oauth-flow");
 async function generatePKCE() {
   const codeVerifier = oauth.generateRandomCodeVerifier();