Minor tweaks to the default ollama client
- switching to alt buffer
- markdown post-processing
- output pagination
- full stdio passthrough
- your CTRL+D and CTRL+C should still work
- the generation stream should still be mirrored to the term as its going
./ollamaGlow.sh <model>if you don't specify a model, the script will default to qwen3-coder:30b