Skip to content

Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent API.

Improve judge result console output to reduce noise

Cursor Agent6f392b67751c
diff --git a/browser_use/agent/service.py b/browser_use/agent/service.py
index b01abfb52..12fd57303 100644
--- a/browser_use/agent/service.py
+++ b/browser_use/agent/service.py
@@ -951,10 +951,27 @@ class Agent(Generic[Context, AgentStructuredOutput]):
# Attach judgement to last action result
if self.history.history[-1].result[-1].is_done:
- self.history.history[-1].result[-1].judgement = judgement
+ last_result = self.history.history[-1].result[-1]
+ last_result.judgement = judgement
- # Log the verdict
+ # Get self-reported success
+ self_reported_success = last_result.success
+
+ # Log the verdict based on self-reported success and judge verdict
if judgement:
+ # If both self-reported and judge agree on success, don't log
+ if self_reported_success is True and judgement.verdict is True:
+ return
+
+ # If agent reported success but judge thinks it failed, show warning
+ if self_reported_success is True and judgement.verdict is False:
+ judge_log = '\n⚠️ \033[33mAgent reported success but judge thinks task failed\033[0m\n'
+ if judgement.failure_reason:
+ judge_log += f' Reason: {judgement.failure_reason}\n'
+ self.logger.info(judge_log)
+ return
+
+ # Otherwise, show full judge result
verdict_color = '\033[32m' if judgement.verdict else '\033[31m'
verdict_text = '✅ PASS' if judgement.verdict else '❌ FAIL'
judge_log = f'\n⚖️ {verdict_color}Judge Verdict: {verdict_text}\033[0m\n'