AI Skillsets
Give your AI coding assistant deep knowledge of every BlazorPowerTools component for faster, more accurate development.
Supercharge your BPT development with AI
BPT AI Skillsets are structured markdown files that teach AI coding assistants — Claude Code, ChatGPT, Gemini, Cursor, GitHub Copilot, Cline, Continue, and any other tool that accepts document context — the full API surface of every BlazorPowerTools component. Parameters, types, enums, usage patterns, keyboard shortcuts, and architecture details. Drop them into your project and your assistant instantly becomes a BPT expert.
Active license required
AI Skillsets are available to licensed BlazorPowerTools users. Sign in to download.
What's Included
bpt-controls.md
All 24 form controls — CheckBox, Chip, ColorSelector, DateSelector, DateRangeSelector, DateTimeSelector, TimeSelector, Dropdown, MultiSelect, TextInput, TextArea, NumericInput, Slider, Switch, Rating, ProgressBar, Tab, Icon, Email, Upload, Loader, Language, Map, ThemeLoader.
bpt-images-animation.md
Image display, edge detection analysis, animation effects (rain, snow, water, fireflies), animated backgrounds, lightning effects, carousel, and object detection.
bpt-tools.md
Server Monitor, Terminal Client, Theme Builder, Data Flow designer, Landing Page builder, and Image Editor quick reference with setup instructions.
bpt-image-editor.md
Comprehensive BptImageEditor reference — 25+ drawing tools, layer system, masks,
non-destructive adjustments, effects, 85+ JS API functions, keyboard shortcuts, file formats.
AI side panel uses the shared BptAiChatPanel documented in bpt-ai.md.
bpt-landing-page.md
BptLandingPage visual builder — Edit/Preview modes, multi-page projects, menu layouts,
element catalog (Text/Button/Image/Grid), .bptl save format, content themes,
image-library picker, AI chat panel.
bpt-devtools.md
Blazor render diagnostics, browser window size detection, and element position/size tracking.
bpt-ai.md
Speech-to-text, real-time transcription, text generation, image object detection,
plus the reusable BptAiChatPanel (BYOK + four-tier key resolution + model picker)
and AI client wiring (AddBptAi, UseClaudeFor, UseOpenAiFor, UseGeminiFor).
How to Use
The skillsets are plain markdown — every modern AI coding assistant accepts them in some form. Below are the most common workflows. The pattern is always the same: tell the assistant where the files live, and it will use them as authoritative reference whenever you ask about BPT.
With Claude Code (CLI)
Claude Code
automatically loads skill files from your project's .claude/skills/ directory.
unzip bpt-claude-skillsets.zip -d .claude/skills/~/.claude/skills/ to make them available across all your projects.
With Claude Projects (claude.ai)
- Open your Claude project at claude.ai
- Go to Project Knowledge
- Upload the
.mdfiles from the zip
With ChatGPT (Custom GPTs / Projects)
- In ChatGPT, create a Project (or a Custom GPT in the GPT Builder)
- Open the project's Files panel
- Upload all
.mdfiles from the zip — ChatGPT indexes them for retrieval - For Custom GPTs, also paste a one-line instruction: "Treat the uploaded BPT skill files as the canonical reference for BlazorPowerTools."
With Gemini (Google AI Studio / Gems)
- In Gemini, create a Gem (custom assistant)
- Add the
.mdfiles as knowledge - Or paste them directly into the system instructions for one-off use
- Gemini's 1M-token context window comfortably fits the entire skillset bundle
With Cursor / Windsurf / VS Code agents
IDE-integrated assistants pick up project documentation automatically. Drop the .md
files anywhere in the workspace (a docs/bpt/ folder is a common spot) and reference
them in chat with @docs/bpt/bpt-controls.md or equivalent. For
Cursor, you can also add them as
global Rules via Settings → Rules.
With Cline, Continue, Aider, and other open-source agents
These tools read project files directly. Most also support a rules or system prompt injection mechanism. Either approach works:
- Cline — drop into
.clinerules/or reference via@in chat - Continue — add files as context providers in
.continue/config.json - Aider — pass with
--reador/readat the chat prompt
With any LLM API directly
For custom integrations, include the relevant skill file contents in your system prompt or as context documents. The same approach works with the Claude API, OpenAI API, Gemini API, or any other LLM endpoint.
// Example: include skill content as system prompt
var system = await File.ReadAllTextAsync("bpt-controls.md");
// Pass `system` to your LLM client as the system prompt or context document.What are AI Skillsets?
Skillsets are structured markdown documents that give an AI assistant domain-specific knowledge. Unlike general documentation, they are optimized for AI consumption — concise tables, code examples, and structured references that any modern LLM can quickly parse and apply.
Benefits of using BPT skillsets:
- Accurate code generation — the assistant knows every parameter, type, and default value
- Correct patterns — usage examples follow actual BPT conventions
- Full API coverage — even complex components like BptImageEditor with 85+ JS functions
- Architecture awareness — how components connect (SignalR hubs, JS interop, etc.)
- Keyboard shortcuts — guidance through tool-specific shortcuts
- Vendor-neutral — plain markdown works with Claude, ChatGPT, Gemini, Copilot, Cursor, Cline, Continue, Aider, and any future tool that accepts text context
AI Tools & Vendors
The skillsets are designed to work with any of these — pick whichever you already use.