Firehose for 28 March 2025

Two desktop and scripting tools for local LLM work: an app for running batch ideation prompts against a local model, and a Python script that bulk updates model entries in Open WebUI by swapping the base model ID.