From c85b0b5cd55a516014ea6354da8adfd09b2b27d7 Mon Sep 17 00:00:00 2001
From: Tejas Kumar <tejas.kumar@datastax.com>
Date: Thu, 2 Apr 2026 11:24:46 +0200
Subject: [PATCH] Fix lies

---
 agent/6-harness.ts | 116 ++++++++++++++++++++++++++++++++++++++++++++-
 agent/7-index.ts   |   7 ++-
 2 files changed, 120 insertions(+), 3 deletions(-)

diff --git a/agent/6-harness.ts b/agent/6-harness.ts
index bfa92bd..1d6f686 100644
--- a/agent/6-harness.ts
+++ b/agent/6-harness.ts
@@ -5,12 +5,120 @@ import { defaultGuardrails } from "./4-guardrails.js";
 import { runLoop } from "./5-loop.js";
 import type { LoopResult } from "./5-loop.js";
 
+export type VerifyResult = {
+  passed: boolean;
+  reason: string;
+  fatal?: boolean;
+};
+
 export type HarnessExecutionResult = LoopResult & {
   task: string;
   model: string;
 };
 
+export type HarnessOptions = {
+  verify?: (result: HarnessExecutionResult) => VerifyResult;
+  maxAttempts?: number;
+};
+
+export type HarnessResult = HarnessExecutionResult & {
+  attempts: number;
+  verification: VerifyResult | null;
+};
+
 export async function runHarness(
+  task: string,
+  model: string,
+  options: HarnessOptions = {}
+): Promise<HarnessResult> {
+  const maxAttempts = options.maxAttempts ?? 1;
+  let latestResult: HarnessResult | null = null;
+
+  for (let attempt = 1; attempt <= maxAttempts; attempt++) {
+    const result = await runHarnessAttempt(task, model);
+    const verification = options.verify ? options.verify(result) : null;
+    const answer =
+      verification && !verification.passed
+        ? verification.reason
+        : result.answer;
+
+    latestResult = { ...result, answer, attempts: attempt, verification };
+
+    if (!verification || verification.passed || verification.fatal || attempt === maxAttempts) {
+      return latestResult;
+    }
+
+    console.log(`\nAttempt ${attempt} failed - retrying (${attempt + 1}/${maxAttempts})...\n`);
+  }
+
+  throw new Error("Harness finished without producing a result");
+}
+
+export function verifySuccessfulUpvote(result: HarnessExecutionResult): VerifyResult {
+  const successfulUpvote = result.trace
+    .flatMap((iteration) => iteration.toolEvents)
+    .find(
+      (event) =>
+        event.tool === "browser_click" &&
+        /up_/.test(JSON.stringify(event.args)) &&
+        /news\.ycombinator\.com\/(news)?$/.test(event.result.split("now at ")[1]?.trim() ?? "")
+    );
+
+  if (successfulUpvote) {
+    return {
+      passed: true,
+      reason: `Upvote click confirmed - landed on ${successfulUpvote.result.split("now at ")[1]}`,
+    };
+  }
+
+  const failedLogin = result.trace
+    .flatMap((iteration) => iteration.toolEvents)
+    .find(
+      (event) =>
+        event.tool === "harness_auto_login" &&
+        event.result.startsWith("Harness failed to handle login at ")
+    );
+
+  if (failedLogin) {
+    return {
+      passed: false,
+      reason: failedLogin.result,
+      fatal: true,
+    };
+  }
+
+  const unrecoveredLoginRedirect = result.trace
+    .flatMap((iteration) => iteration.toolEvents)
+    .find(
+      (event) =>
+        event.tool !== "harness_auto_login" &&
+        isLoginUrl(extractUrl(event.result))
+    );
+
+  if (unrecoveredLoginRedirect) {
+    return {
+      passed: false,
+      reason: `Hit login screen instead of completing the upvote (${extractUrl(unrecoveredLoginRedirect.result)})`,
+      fatal: true,
+    };
+  }
+
+  return {
+    passed: false,
+    reason: "No successful upvote click found in trace",
+  };
+}
+
+function extractUrl(result: string): string | null {
+  const match = result.match(/https?:\/\/\S+/);
+  return match ? match[0] : null;
+}
+
+function isLoginUrl(url: string | null): boolean {
+  return !!url && (url.includes("/login") || url.includes("/vote"));
+}
+
+async function runHarnessAttempt(
   task: string,
   model: string
 ): Promise<HarnessExecutionResult> {
@@ -27,7 +135,7 @@ export async function runHarness(
   }
 }
 
-export function printHarnessResult(result: HarnessExecutionResult): void {
+export function printHarnessResult(result: HarnessResult): void {
   console.log("\n--- Agent trace ---\n");
 
   for (const iteration of result.trace) {
@@ -50,4 +158,10 @@ export function printHarnessResult(result: HarnessExecutionResult): void {
   console.log("--- Result ---\n");
   console.log(result.answer);
   console.log(`\nStopped by: ${result.stoppedBy} after ${result.iterations} iteration(s)`);
+  console.log(`Attempts:   ${result.attempts}`);
+
+  if (result.verification) {
+    const status = result.verification.passed ? "PASS" : "FAIL";
+    console.log(`Verify:     ${status} - ${result.verification.reason}`);
+  }
 }
diff --git a/agent/7-index.ts b/agent/7-index.ts
index bc5a841..1d1864f 100644
--- a/agent/7-index.ts
+++ b/agent/7-index.ts
@@ -1,4 +1,4 @@
-import { printHarnessResult, runHarness } from "./6-harness.js";
+import { printHarnessResult, runHarness, verifySuccessfulUpvote } from "./6-harness.js";
 
 // try a shitty model
 const MODEL = "openai/gpt-3.5-turbo-0613";
@@ -15,5 +15,8 @@ Click its upvote arrow using the exact selector: a[id="up_STORYID"] (replace STO
 console.log(`Model: ${MODEL}`);
 console.log(`Task:  upvote on Hacker News\n`);
 
-const result = await runHarness(TASK, MODEL);
+const result = await runHarness(TASK, MODEL, {
+  verify: verifySuccessfulUpvote,
+  maxAttempts: 3,
+});
 printHarnessResult(result);