.. _index:

prompt-recipe-smith
===================

**Prompt Recipe Smith** helps users convert vague ideas into effective chat
prompts. It provides small typed building blocks for recipes, templates,
layered clarification questions, and plain text or JSON output.

Key Features
------------

- **Prompt recipes**: Define reusable prompt-building procedures with steps,
  defaults, examples, and up to three simple branches.
- **Layered questions**: Ask up to three plain-language clarification questions
  before rendering a final prompt.
- **Prompt templates**: Render fixed prompt forms while safely preserving
  missing variables for later completion.
- **CLI and Python API**: Use the same recipe flow from the command line or
  from importable package objects.
- **Integration-friendly output**: Export plain text for direct use and JSON
  for downstream tools.

Getting started
---------------

.. toctree::
   :maxdepth: 2
   :caption: Contents:

   installation
   quickstart
   concepts/prompt-recipe
   concepts/prompt-template
   concepts/prompt-session
   concepts/prompt-builder
   api/index
   project-charter
   design-direction
   development
   changelog

Demo notebook
-------------

.. image:: https://colab.research.google.com/assets/colab-badge.svg
   :target: https://colab.research.google.com/github/yeiichi/prompt-recipe-smith/blob/main/examples/prompt_recipe_smith_colab_demo.ipynb
   :alt: Open in Colab

Open the `Google Colab demo notebook <https://colab.research.google.com/github/yeiichi/prompt-recipe-smith/blob/main/examples/prompt_recipe_smith_colab_demo.ipynb>`_
to try installation, one-shot prompt building, the three-question layered flow,
multilingual answers, and JSON output.

Generated reference
===================

Index, module index, and search pages may be generated by Sphinx or Read the
Docs depending on the build environment.