ranade-oss makes jurisdiction and model choice stick

This fork tackles two familiar sources of legal-AI friction: losing matter context and working from an outdated model menu.

workflowintegration

A lawyer's chosen jurisdiction now persists within the relevant project as they move from an initial question into a live conversation. It also stays in sync across browser tabs, while keeping separate matters separate. The current choices cover Ontario, Canada and the US.

The team has also replaced a fixed catalogue of AI models with live discovery from supported providers.

  • Each provider's available models are refreshed from that provider, with checks for whether a model can actually generate a response.
  • New provider models can appear without someone manually updating a static list.

That is a practical improvement, though the live provider connections deserve a real-world smoke test before anyone relies on them in production.

So what Legal teams running work across matters or evaluating several AI providers will want to see whether this makes model selection less brittle.

View this fork on GitHub →

Spotted something wrong? Or know the PR text has fresher detail than the writeup above?

Commits in this thread

1 commit from ranade-oss/ROSS-RanadeOSS, oldest first. Source extracted verbatim from the harvested git log.

SHA Subject Author Date
58aa0b08 Persist jurisdiction and discover provider models (#84) ranade-oss 2026-08-04 ↗ GitHub
* Persist jurisdiction and discover provider models

* Fix jurisdiction persistence lint

---------

Co-authored-by: Codex <codex@openai.com>

Capture this thread into my fork

Download a single Markdown prompt that tells Claude how to port every commit above into your working tree — adapting paths and structure to match your repo. Run it via claude -p < capture-thread-904.md from inside the repo you want the changes in.

⬇ Download capture-thread-904.md