A local AI workspace with LibreChat and Ollama
A local AI setup narrowed to a concrete Docker-to-host networking problem: the model answered directly, while the container path needed a firewall change.
Explore the mess / 5 entries
Projects, experiments, and notes involving local ai. Follow a topic across otherwise unrelated rabbit holes.
Nothing matches those filters.
Try a broader term, or follow a topic link in the archive.
Browse the notebook →A local AI setup narrowed to a concrete Docker-to-host networking problem: the model answered directly, while the container path needed a firewall change.
Exploring a self-hosted AI workspace led to authorization hardening, blocking CI, local release checks, and a first-run setup test that exposed a missing-directory problem.
A personal lab for watching how an ordinary sudo-capable agent handled normal Linux tasks, including malformed actions and a suspected loop.
A local AI desktop concept developed around Linux identities, a trusted broker, staged outputs, and evidence that could actually support its security claims.
From a short image-to-video reel to a basic text-to-image graph, with the actual nodes, settings, and failures that shaped the experiment.