# Setup Guide — Handbook Generator (Streamlit + Supabase + Gemini)

## Prerequisites
- Python 3.10+ installed
- A Google Gemini API key
- A Supabase project (Postgres) with pgvector enabled

---

## 1) Clone repo + create virtual environment
```bash
git clone
cd handbook-generator

python -m venv venv
# Windows:
venv\Scripts\activate
# Mac/Linux:
# source venv/bin/activate

2) Install dependencies

pip install -r requirements.txt

---

3) Supabase setup (one-time)

3.1 Create Supabase project

Create a project in Supabase, then copy:

• Project URL
• anon public key (from: Project Settings → API)

3.2 Enable pgvector

In Supabase → SQL Editor, run:

create extension if not exists vector;

3.3 Create documents table

(Embedding dimension is 3072 for gemini-embedding-001)

create table if not exists documents (
  id bigserial primary key,
  content text not null,
  metadata jsonb,
  embedding vector(3072)
);

3.4 Create similarity search function

create or replace function match_documents(
  query_embedding vector(3072),
  match_count int default 5
)
returns table(id bigint, content text, similarity float)
language sql stable
as $$
  select
    documents.id,
    documents.content,
    1 - (documents.embedding <=> query_embedding) as similarity
  from documents
  order by documents.embedding <=> query_embedding
  limit match_count;
$$;

---

4) Set environment variables

Windows (PowerShell)

setx GEMINI_API_KEY "YOUR_GEMINI_KEY"
setx SUPABASE_URL "YOUR_SUPABASE_PROJECT_URL"
setx SUPABASE_ANON_KEY "YOUR_SUPABASE_ANON_KEY"

Close and re-open your terminal after running setx.

Mac/Linux

export GEMINI_API_KEY="YOUR_GEMINI_KEY"
export SUPABASE_URL="YOUR_SUPABASE_PROJECT_URL"
export SUPABASE_ANON_KEY="YOUR_SUPABASE_ANON_KEY"

---

5) Run the app

streamlit run app.py

Open the local URL Streamlit prints (usually http://localhost:8501).

---

6) How to use (quick test)

1. Upload a PDF
2. Click Index this PDF in Supabase
3. Ask a question in the chat input → verify contextual answer
4. Enter a handbook topic → click Generate Handbook → verify output reaches 20,000+ words
5. Download the generated handbook.md

---

Troubleshooting

• If Gemini model name fails: use gemini-2.5-flash for generation.
• If search returns nothing: make sure you indexed at least one PDF and the SQL function exists.
• If Supabase insert fails: verify vector extension is enabled and embedding dimension is 3072.

If you want, paste your current `requirements.txt` here and I’ll clean it up (remove extras) so your setup guide matches exactly.