TuxNest · local-first AI engineering
Troubleshooting
When TuxNest behaves unexpectedly, verify the runtime, model capability, context and tool execution independently.
Built around control.
Minimal interface, explicit model choice, reviewable changes and evidence you can inspect.
01Provider unavailable
Confirm the configured endpoint is reachable.
02Model missing
Check the model exists in the selected runtime.
03Tools absent
Verify capability detection and tool selection.
04Weak repository answer
Increase evidence coverage and inspect actual files.
05Remote GPU issue
Verify SSH and Ollama independently before debugging the extension.
Local AI should feel local.
Your editor, your model runtime, your hardware and your approval remain first-class parts of the workflow.