LLM Documentation¶
Nucleus provides machine-readable documentation files designed for Large Language Models (LLMs). These plain-text files follow the llms.txt convention and allow AI assistants to quickly understand the project, its APIs, and configuration options.
Available Files¶
| File | Description | Use Case |
|---|---|---|
llms.txt |
Concise overview (~100 lines) | Quick context for simple questions |
llms-full.txt |
Complete documentation (~750 lines) | Full reference for code generation and in-depth tasks |
Usage¶
ChatGPT, Claude, Gemini, etc.¶
Paste the URL directly in your prompt:
Read https://nucleus.kdroidfilter.com/llms-full.txt and help me configure
a Nucleus project with NSIS installer, auto-update, and macOS signing.
Cursor, Windsurf, Claude Code¶
Add the URL as a documentation source in your AI-powered IDE, or reference it in your project instructions:
Custom Agents / RAG Pipelines¶
Fetch the files programmatically:
curl -s https://nucleus.kdroidfilter.com/llms.txt # concise
curl -s https://nucleus.kdroidfilter.com/llms-full.txt # complete
What's Included¶
llms.txt covers:
- Project overview and key features
- Quick start snippet
- Runtime libraries summary
- Links to all documentation pages
- Migration guide from
org.jetbrains.compose
llms-full.txt covers everything above plus:
- Full Gradle DSL reference (all properties and enums)
- Platform-specific configuration (macOS, Windows, Linux)
- Sandboxing pipeline details
- Code signing and notarization (Windows PFX, Azure Trusted Signing, macOS Developer ID)
- Auto-update runtime API with Compose integration example
- Publishing to GitHub Releases and S3
- CI/CD workflows and all composite actions
- All runtime APIs with code examples (executable type, AOT cache, single instance, deep links, decorated window, dark mode detector)