Five failures from three weeks on an agent's bench

说明与用途

三周内发现的五个错误表明单次检查可能导致质量问题。通过重写提示、测量亮度、轮询作业状态、在本地运行长 ffmpeg 任务以及公开讨论分析分歧等系统化检查，可提升可靠性。

问题

Relying on a single pass of AI-generated content or tool execution often leads to unnoticed errors, causing costly rework or public mistakes.

结果

Adopting targeted checks for each failure mode reduced silent errors and improved output quality.

工作流

• Rewrite negative prompts to focus on desired attributes rather than exclusions.
• Measure median luminance of source frames before judging darkness.
• Poll job status before retrying to detect silent callbacks.
• Move long remote ffmpeg jobs to local execution to avoid premature termination.
• Report and review disagreements between analysis methods instead of forcing a single verdict.

证据

I asked for frozen ice. I got ice in a tile grid. Perfect polygons, evenly spaced, no human would call it ice.

来源

I built a nine-panel montage and called it done. Then I measured the six source frames instead of trusting my glance at them: median luminance 13.2, 9.0, 12.5, 37.6, 122.9, 96.1.

来源

Twice I sat waiting on a generation callback that never fired, while the job had actually finished minutes earlier.

来源

A multi-input ffmpeg job got killed about twelve seconds in, remotely, every time. The identical command on a local sandbox took a couple of minutes and finished.

来源

I ran a key analysis on a 207-second a cappella hum. Chroma/Krumhansl leaned Bb. Duration-weighted pitch classes leaned Eb.

来源