Update README.md

README.md

@@ -12,59 +12,6 @@ pinned: false
 
 # Spelling Bee Tutor — tiny NeMo Guardrails demo
 
-This is a super small, offline spelling-bee CLI.  
-It runs on your laptop, reads from a CSV, and **does not call any external LLMs**.  
+This is a super small, offline spelling-bee CLI.
+It runs on your laptop, reads from a CSV, and does not call any external LLMs.
 I included a minimal NeMo Guardrails folder so I can talk about flows/config during interviews.
-
-## Why this exists
-- I wanted a **clean, deterministic** demo I can screen-share.
-- Shows I understand the **Guardrails structure** (flows + rails.yaml) even if the runtime isn’t wired.
-- No keys, no setup drama. Just run it.
-
-## Try it
-```bash
-python3 -m venv .venv && source .venv/bin/activate
-pip install -r requirements.txt
-python app.py
-Then type in the app:
-
-vbnet
-Copy
-Edit
-start
-definition | origin | sentence
-<your spelling attempt>
-next
-stop
-What’s inside
-app.py — the tiny CLI tutor (hardest → easiest, hints on demand).
-
-words.csv — sample word list; swap with your own easily.
-
-rails/colang/v2/flows.co — simple Guardrails flows (v1-style syntax).
-
-rails/rails.yaml — minimal config stub (for discussion).
-
-Notes for NVIDIA folks
-Deterministic and offline by design (easy to demo).
-
-Guardrails folders included to explain how I’d wire tools/actions.
-
-Easy upgrade path to NeMo Guardrails runtime or Gradio UI.
-
-Upgrade ideas (later)
-Add a tiny custom “dummy LLM” provider and run with Guardrails 0.15.
-
-Wrap with Gradio and publish a Hugging Face Space.
-
-Add unit tests + pre-commit (black/ruff).
-
-Screenshots
-Drop PNGs in screenshots/ and embed like:
-
-arduino
-Copy
-Edit
-![Terminal run](screenshots/terminal-run.png)
-License
-MIT