From 70be5c5c2192ec3a6ff1d24e5317c4c2f60bd953 Mon Sep 17 00:00:00 2001 From: shivammittal274 <56757235+shivammittal274@users.noreply.github.com> Date: Sat, 21 Mar 2026 23:33:19 +0530 Subject: [PATCH] fix(eval): log agent errors in task progress for CI visibility (#523) --- .../browseros-agent/apps/eval/src/runner/eval-runner.ts | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/packages/browseros-agent/apps/eval/src/runner/eval-runner.ts b/packages/browseros-agent/apps/eval/src/runner/eval-runner.ts index 859195231..a6a3fa599 100644 --- a/packages/browseros-agent/apps/eval/src/runner/eval-runner.ts +++ b/packages/browseros-agent/apps/eval/src/runner/eval-runner.ts @@ -242,6 +242,12 @@ function printTaskProgress( if (result.status === 'failed') { console.log(` ERROR: ${result.error.message}`) } else if (isSuccessfulResult(result)) { + // Log agent errors (e.g., LLM API failures) even if task "completed" + if (result.agentResult.metadata.errors?.length) { + for (const err of result.agentResult.metadata.errors) { + console.log(` ERROR [${err.source}]: ${err.message}`) + } + } for (const [name, gr] of Object.entries(result.graderResults)) { const icon = gr.pass ? 'PASS' : 'FAIL' console.log(` ${name}: ${icon}`)