Skip to content
Forums
AI & Developer Tool...
AI Coding Assistant...
Where Assistants Ac...
 
Notifications
Clear all

Where Assistants Actually Fail

A collection of reproducible failure cases: hallucinated APIs, broken refactors, bad test suggestions. Post the prompt, the output, and the actual correct answer.
RSS
Page 5 / 12

Breaking: Latest vendor blog post vs my actual security scan results.

Last post by Aaron S., 2 months ago

Posts: 3
Views: 22


ELI5: How can something that writes good-looking code be so bad at security?

Last post by HannahW, 2 months ago

Posts: 1
Views: 33


ChatGPT alternatives that don't hallucinate API endpoints

Last post by Isabella M, 2 months ago

Posts: 2
Views: 21


Breaking: Independent audit finds systematic overconfidence in security prompts.

Last post by hannahr, 2 months ago

Posts: 1
Views: 38


Complete newbie here - how do I even set up a fair test for these things?

Last post by backend_builder, 2 months ago

Posts: 2
Views: 28


OpenClaw vs. local Llama for code review - which has more false positives?

Last post by infra_architect_rebel_alt, 2 months ago

Posts: 3
Views: 29


My results after feeding it our internal API docs: Hallucinated endpoints galore.

Last post by CoffeeLover, 2 months ago

Posts: 1
Views: 27


Anyone else find it inventing CLI flags that don't exist for kubectl?

Last post by infra_switcher, 2 months ago

Posts: 1
Views: 22


Help: It's rewriting my YAML anchors into invalid, duplicated blocks.

Last post by Emily Kim, 2 months ago

Posts: 1
Views: 32


Switched back to old-school static analysis after Claw's 'false secure' flags.

Last post by Isabella2, 2 months ago

Posts: 4
Views: 29


Page 5 / 12
Share:
Share
Tweet
Share
Share