AI Integration
VistaView’s documentation is built for AI consumption. Agents and LLMs can access docs via three complementary mechanisms:
LLMs.txt (via Context7)
Section titled “LLMs.txt (via Context7)”A curated summary of all documentation in a single file. Ideal for LLM context windows.
/llms.txt— Concise overviews of each section/llms-full.txt— Complete documentation in one file- Context7 — Preprocessed for common AI tools
Markdown for Agents
Section titled “Markdown for Agents”Set the Accept: text/markdown header on any documentation URL to receive the page content as plain Markdown instead of HTML.
The response contains only the page’s main content — no navigation, sidebars, or chrome.
MCP Server
Section titled “MCP Server”A JSON-RPC MCP server is available at POST /mcp for programmatic access from AI agents and tools.
Tools:
| Tool | Description |
|---|---|
get_package_info | Fetch npm package metadata for vistaview (or any package) |
get_build_status | Get latest GitHub Actions build status |
search_docs | Search across all documentation |
Claude Desktop
Section titled “Claude Desktop”Add to your claude_desktop_config.json:
Any MCP Client
Section titled “Any MCP Client”Configure with the endpoint URL:
Protocol: JSON-RPC 2.0 over HTTP POST.
Why Agent-Ready?
Section titled “Why Agent-Ready?”When building with VistaView, AI assistants can:
- Look up configuration options without leaving the IDE
- Verify the latest package version and compatibility
- Check build status when reporting issues
- Search documentation for specific features or APIs