Update the main library documentation to better describe the document
engine as LLM-guided from indexing to querying, emphasizing the
hierarchical semantic tree approach without vector databases or
similarity search.

- Add .rustfmt.toml with proper formatting configuration including
  import granularity, grouping, and max width settings
- Simplify imports in client/builder.rs using unified import syntax
- Update EngineBuilder documentation with clearer examples
- Make endpoint requirement explicit in builder documentation
- Remove unused ConfigLoader import from client builder

- Remove the `config` field from EngineBuilder struct as it was unused
- Remove the `with_config` method since it's no longer needed
- Simplify the build process by directly using Config::default()
- Remove Config-related error variant from BuildError enum
- Update all related code to work without custom config field

- Remove unused RetrievalConfig import and field from EngineBuilder
- Delete with_retrieval_config method that was no longer needed
- Clean up related configuration application logic in build method

- Remove unused imports including MemoStore, reorder imports following
  crate granularity rules
- Remove top_k, fast_mode, precise_mode, and memo_store fields from
  EngineBuilder as they are no longer needed
- Remove related methods with_top_k, fast, precise, and with_memo_store
- Clean up build logic to remove configuration of removed features
- Delete associated test cases for removed functionality
- Remove rustfmt.toml file and advanced example that are no longer used

- Add optional Config field to EngineBuilder struct for advanced tuning
- Implement with_config method to set custom configuration
- Use provided config or default when building engine
- Export Config struct from main library module

The new configuration option allows users to provide custom settings
for advanced parameter tuning while maintaining backward compatibility
through builder methods that can still override individual fields.

Add comprehensive Python wrapper classes for the vectorless engine,
including Config, IndexContext, QueryContext, DocumentInfo, Engine,
and DocumentGraph with full async support and proper error handling.

The Python bindings provide complete access to the core functionality:
- Engine class with async index/query operations
- Config for advanced engine tuning
- Context classes for indexing and querying
- Document graph for cross-document relationships
- Proper error handling with VectorlessError
- Full async/await support using pyo3-async-runtimes

- Add validation to check if LLM endpoint is configured
- Return MissingEndpoint error when endpoint is not provided
- Update BuildError enum with MissingEndpoint variant
- Include proper error message for missing endpoint configuration

- Remove unnecessary line breaks in import statements in context.rs
- Format long import statements with proper line breaks in lib.rs
- Fix multiline string formatting in builder.rs error definition
- Add missing endpoint example in engine.rs documentation
- Reorganize import statements in engine.rs for better readability

- remove unused FaStar import from react-icons/fa
- adjust GitHub star button styling with larger icon size (14px to 16px)
- update spinner character from three dots to HTML entity …
- fix stargazers link to point to correct repository path
- modify font weights in styles module from 700 to 600 for consistency
- redesign homepage layout with new typography and spacing
- add dark theme support for all components
- implement responsive design improvements
- update navbar styling with new dimensions and colors
- replace hero section content with emphasis on reasoning approach
- enhance GitHub star button with improved styling and SVG icons
- update CSS variables for better color scheme consistency

BREAKING CHANGE: homepage layout structure has been completely redesigned

- Remove old feature grid component and styling
- Add new demo card component with dark theme design
- Implement tabbed interface for Python/Rust code examples
- Add syntax highlighting for Python and Rust code blocks
- Include copy-to-clipboard functionality for code snippets
- Create terminal output simulation with blinking cursor animation
- Add installation command section with copy button
- Update homepage to use new get started section instead of features

- replace step-by-step text explanation with visual workflow diagram
- remove numbered steps component and associated CSS styles
- add new workflow wrapper and image styling
- update section title to "How does Vectorless work?"
- add subtitle "You declare a few lines of code. We do everything else."
- use SVG image for better visual representation of the process

- Implement responsive slider component with smooth transitions and navigation controls
- Add 6 detailed use cases showcasing financial reports, legal documents,
  technical documentation, research papers, cross-document analysis,
  and compliance scenarios
- Create CSS module styles for slider track, cards, navigation buttons,
  and dot indicators with active states
- Add keyboard-friendly navigation and screen reader accessibility
- Include realistic example queries and answers for each use case
- Make slider responsive with different card sizes for various screen widths
- Add resize event listener to handle dynamic layout adjustments

- Add new dark themed CTA section with class sectionCtaDark
- Implement installation command cards with copy functionality for pip and cargo
- Add interactive copy buttons with success feedback animations
- Update CTA title to "Start reasoning, not vectoring"
- Style CTA elements with new typography and spacing
- Add GitHub star button with custom styling
- Implement responsive design for installation cards

- Increase footer padding from 3rem 1.5rem 2rem to 5rem 1.5rem 3rem
- Adjust font sizes for footer title (0.8rem -> 0.85rem), link items
  (0.88rem -> 0.92rem), and copyright text (0.8rem -> 0.82rem)
- Update line height for footer links from 1.8 to 2
- Increase margin-bottom for footer title from 0.75rem to 1rem
- Adjust copyright section margins and padding (2rem -> 3rem top margin,
  1.5rem -> 2rem padding-top)

style(landing): update gradient color scheme

- Change hero section gradient from light colors (#EBEDF0, #AF788B)
  to darker palette (#1e293b, #8B5E6F)

refactor(landing): remove dark section class from CTA component

- Remove styles.sectionCtaDark class from section element in SectionCTA
  component

- Replace copyright symbol from literal "©" to Unicode escape "\u00A9"
- Reduce footer bottom padding from 3rem to 1.5rem
- Adjust footer copyright styling: increase font size from 0.82rem to 0.88rem,
  remove center alignment, letter spacing, top margin, and border

- Replace the technical explanation of semantic tree navigation
with clearer value proposition highlighting supported document
types (PDFs, Markdown, reports, contracts)
- Rename "Why Vectorless" section to "How It Works" for better
clarity
- Remove detailed technical explanation about semantic tree
indexing approach

- Replace gradient text effect with standard CSS variable
- Remove unnecessary background properties for better maintainability
- Use consistent text color through --text variable

- Add with_endpoint method to EngineBuilder in Rust implementation
- Include endpoint parameter in Python Engine constructor
- Update documentation examples to show endpoint configuration
- Set default OpenAI endpoint to https://api.openai.com/v1

fix(docs): update hero title emphasis color to use text variable

- Change color from --primary-dark to --text in index module CSS
- Maintain consistent styling with the rest of the documentation

…kage

- Update workspace package version from 0.1.28 to 0.1.29 in Cargo.toml
- Update python package version from 0.1.7 to 0.1.8 in pyproject.toml