Add Flask prompts-bot scaffold and prompts index

[jasperphipps]

Why

Provide a lightweight HTTP API to browse and search the repository's prompts so the content library can be used as a local prompts-serving bot.

What it does

• Adds a minimal Flask app under bot/ with endpoints:
  • GET /health
  • GET /prompts?q=&category=
  • GET /prompts/
• Includes requirements.txt, PROMPTS_BOT_README.md (run instructions), .gitignore
• Adds the generated zetalib_prompts_index.csv and zetalib_prompts_index.json used by the service

Approach

The Flask app loads zetalib_prompts_index.json at startup and performs simple substring searches over path, excerpt, and category. The /prompts/ endpoint returns the prompt file content for convenience. This is intentionally small and easy to run locally.

Notes and considerations

• For production use: add pagination, input sanitization, rate limiting, authentication, and caching.
• The index file is generated by tooling; re-run the indexer if prompts change.
• Origin was not writable from this environment; the PR creation flow will create/push to an appropriate fork if required.

Testing

1. python3 -m venv .venv && source .venv/bin/activate
2. pip install -r requirements.txt
3. export FLASK_APP=bot.app
4. flask run --host=0.0.0.0 --port=5000
5. Visit /prompts and /prompts?q=your+term