-
Visually distinguish the different components
-
Add guardrails to limit token usage
-
Make settings saved in local storage
-
Edit OpenAPI key via the UI
-
Host a demo version with pre-populated data on GitHub pages
-
Optimize prompting to reduce token usage. Use another cheaper LLM to do ID filtering, and one for query generation?
-
Redux-ify ChatGPT API class?
-
Redux-ify runQuery context?