Tag
“error handling”
2 posts
- Product & Mission··6 min read
Why We Made Our AI Tools Fail Loudly Instead of Returning Your Original Image
A tool that quietly returns your input when the model fails looks like it worked — and wastes your time worse than an error would. The "no silent fallback" rule, and the three real bugs it caught in our own code.
- Technical Deep Dives··5 min read
"Session Has No Input Names": The Anatomy of a Truncated Model Download
ONNX Runtime doesn't throw when you hand it a half-downloaded model. It builds a broken session that fails cryptically three tools later. The failure chain and the two checks that make it fail at the source.