Add API Documentation with MkDocs

[lmz]

Problem:
The project has some Python docstrings, type hints, and well-documented data models, but no centralized API documentation. Additionally, there is substantial documentation within a variety of Readme markdowns which include documentation about input, output and algorithms. This creates a parallel set of documentation alongside the doe, which creates a barrier to keeping the documentation up-to-date.

Proposed Solution

• Move the Readme documentation to the python docstrings alongside code, which makes it easier to keep up-to-date and where it sits alongside argument documentation.
• Generate API documentation using MkDocs Material which will generate documentation from docstrings, with automatic deployment to GitHub Pages.

Documentation Coverage:

• Codebook: All enumeration types (trip purposes, modes, person types, etc.)
• Data Models: Pydantic models for survey data and model outputs (DaySim, CTRAMP)
• Pipeline: Pipeline class, caching mechanism, step decorators
• Pipeline Steps: All @step() decorated functions with algorithm details

Technical Approach:

• MkDocs Material theme for clean, searchable interface
• mkdocstrings[python] for auto-generation from docstrings
• griffe-pydantic extension for proper Pydantic model documentation
• GitHub Actions workflow for automatic deployment on merge to main
• Strict mode to catch broken cross-references