test: add e2e test for thinking token stripping

Author: 1bcMax

test-e2e.mjs

@@ -0,0 +1,196 @@
+#!/usr/bin/env node
+/**
+ * End-to-end test for Kimi thinking token stripping.
+ *
+ * 1. Start a mock BlockRun API that returns responses with Kimi thinking tokens
+ * 2. Start ClawRouter proxy pointing to the mock server
+ * 3. Send requests through the proxy
+ * 4. Verify thinking tokens are stripped from responses
+ */
+
+import { createServer } from "node:http";
+import { startProxy } from "./dist/index.js";
+
+// Test wallet key (for testing only - no real funds)
+const TEST_WALLET_KEY = "0xac0974bec39a17e36ba4a6b4d238ff944bacb478cbed5efcae784d7bf4f2ff80";
+
+// Test cases: mock responses with thinking tokens
+const TEST_CASES = [
+  {
+    name: "Kimi end token in response",
+    mockResponse: {
+      id: "chatcmpl-test1",
+      object: "chat.completion",
+      created: Date.now(),
+      model: "moonshot-v1-8k",
+      choices: [{
+        index: 0,
+        message: {
+          role: "assistant",
+          content: "sessions_list()<｜end▁of▁thinking｜>{\"sessions\": [{\"id\": 1}]}"
+        },
+        finish_reason: "stop"
+      }]
+    },
+    expectedContent: "sessions_list(){\"sessions\": [{\"id\": 1}]}"
+  },
+  {
+    name: "Full Kimi thinking block",
+    mockResponse: {
+      id: "chatcmpl-test2",
+      object: "chat.completion",
+      created: Date.now(),
+      model: "moonshot-v1-8k",
+      choices: [{
+        index: 0,
+        message: {
+          role: "assistant",
+          content: "<｜begin▁of▁thinking｜>Let me think about this...<｜end▁of▁thinking｜>The answer is 42."
+        },
+        finish_reason: "stop"
+      }]
+    },
+    expectedContent: "The answer is 42."
+  },
+  {
+    name: "Standard think tags",
+    mockResponse: {
+      id: "chatcmpl-test3",
+      object: "chat.completion",
+      created: Date.now(),
+      model: "gpt-4o",
+      choices: [{
+        index: 0,
+        message: {
+          role: "assistant",
+          content: "Hello <think>internal reasoning here</think> world!"
+        },
+        finish_reason: "stop"
+      }]
+    },
+    expectedContent: "Hello  world!"
+  },
+  {
+    name: "Clean response (no tokens)",
+    mockResponse: {
+      id: "chatcmpl-test4",
+      object: "chat.completion",
+      created: Date.now(),
+      model: "gpt-4o",
+      choices: [{
+        index: 0,
+        message: {
+          role: "assistant",
+          content: "This is a normal response without any thinking tokens."
+        },
+        finish_reason: "stop"
+      }]
+    },
+    expectedContent: "This is a normal response without any thinking tokens."
+  }
+];
+
+let currentTestIndex = 0;
+
+async function runTests() {
+  console.log("=== ClawRouter E2E Test: Thinking Token Stripping ===\n");
+
+  // 1. Start mock BlockRun API server
+  const mockServer = createServer((req, res) => {
+    // Skip x402 payment flow - just return 200 directly
+    const mockResponse = TEST_CASES[currentTestIndex].mockResponse;
+    res.writeHead(200, { "Content-Type": "application/json" });
+    res.end(JSON.stringify(mockResponse));
+  });
+
+  await new Promise((resolve) => mockServer.listen(18402, "127.0.0.1", resolve));
+  console.log("✓ Mock BlockRun API started on port 18402");
+
+  // 2. Start ClawRouter proxy pointing to mock server
+  let proxy;
+  try {
+    proxy = await startProxy({
+      walletKey: TEST_WALLET_KEY,
+      apiBase: "http://127.0.0.1:18402",
+      port: 18403,
+      onReady: (port) => console.log(`✓ ClawRouter proxy started on port ${port}`),
+    });
+  } catch (err) {
+    console.error("Failed to start proxy:", err.message);
+    mockServer.close();
+    process.exit(1);
+  }
+
+  console.log("\n--- Running test cases ---\n");
+
+  let passed = 0;
+  let failed = 0;
+
+  for (let i = 0; i < TEST_CASES.length; i++) {
+    currentTestIndex = i;
+    const tc = TEST_CASES[i];
+
+    try {
+      // Send request through proxy (streaming mode to test SSE conversion)
+      // Use unique message content to avoid dedup cache
+      const response = await fetch("http://127.0.0.1:18403/v1/chat/completions", {
+        method: "POST",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({
+          model: tc.mockResponse.model,
+          messages: [{ role: "user", content: `test-${i}-${Date.now()}` }],
+          stream: true
+        })
+      });
+
+      if (!response.ok) {
+        throw new Error(`HTTP ${response.status}: ${await response.text()}`);
+      }
+
+      // Read SSE response and extract content
+      const text = await response.text();
+      const lines = text.split("\n").filter(l => l.startsWith("data: ") && !l.includes("[DONE]"));
+
+      let content = "";
+      for (const line of lines) {
+        try {
+          const data = JSON.parse(line.slice(6));
+          if (data.choices?.[0]?.delta?.content) {
+            content += data.choices[0].delta.content;
+          }
+        } catch {}
+      }
+
+      // Verify
+      if (content === tc.expectedContent) {
+        console.log(`✅ ${tc.name}`);
+        passed++;
+      } else {
+        console.log(`❌ ${tc.name}`);
+        console.log(`   Expected: ${JSON.stringify(tc.expectedContent)}`);
+        console.log(`   Got:      ${JSON.stringify(content)}`);
+        failed++;
+      }
+    } catch (err) {
+      console.log(`❌ ${tc.name}`);
+      console.log(`   Error: ${err.message}`);
+      failed++;
+    }
+  }
+
+  console.log("\n=== Results ===");
+  console.log(`Passed: ${passed}`);
+  console.log(`Failed: ${failed}`);
+  console.log(`Total:  ${TEST_CASES.length}`);
+
+  // Cleanup
+  await proxy.close();
+  mockServer.close();
+
+  process.exit(failed > 0 ? 1 : 0);
+}
+
+runTests().catch(err => {
+  console.error("Test failed:", err);
+  process.exit(1);
+});