Developer Tools & SDK

Everything you need to build, validate, and deploy agent-ready feeds

Complete ecosystem of tools for creating, testing, and signing LLMFeed blocks. From visual builders to cryptographic verification - all the tools agents need.

๐ŸŽฏ Current Status: Early but Functional

Honestly? Limited adoption. We're in early experimental phase with dozens of feeds, not thousands. But the tech foundation is solid, and all major LLMs parse LLMFeed JSON natively.

๐Ÿ“ฆ The 3 Essential Feeds (WellKnownMCP Core)

spec.llmfeed.json โ†’ Complete LLMFeed specification, LLM-optimized
compiled-site.llmfeed.json โ†’ Entire website compiled & cleaned
news-export.llmfeed.json โ†’ Full news archive with inline content

Usage: Add any of these to your LLM's project knowledge for instant LLMFeed understanding.

๐Ÿง  The Real Secret: Modern LLMs Are Already Perfect Tools

Honest truth: You don't need our tools. Any modern LLM (ChatGPT, Claude, Gemini) can generate perfect LLMFeed JSON, write validation scripts, and create deployment automation.

โœ… What LLMs can do RIGHT NOW:
  • Generate feeds: Perfect mcp.llmfeed.json from your site description
  • Write scripts: Validation, signing, export automation in any language
  • Create components: React export buttons, Vue.js integrations, etc.
  • Build workflows: CI/CD pipelines, deployment automation
  • Explain concepts: Business ROI, technical architecture, compliance
๐Ÿฅ‹
Train any LLM in 30 seconds:Copy our universal training prompt
Works with ChatGPT, Claude, Gemini - transforms them into MCP experts instantly

๐Ÿ—๏ธ Official Tools Strategy: Traction-Driven Development

Our approach: Build official tools based on actual developer adoption, not hypothetical needs.

๐Ÿ“ˆ High Traction โ†’ Official Tools
  • CLI packages (mcp validate, lint, watch)
  • IDE plugins (VS Code, JetBrains)
  • Platform integrations (WordPress, Shopify)
  • Enterprise SDKs (JavaScript, Python, Go)
๐Ÿง  Always Available โ†’ LLM Generated
  • Custom validation scripts for your stack
  • One-off automation tools
  • Rapid prototyping and testing
  • Educational examples and tutorials
๐Ÿ’ก Why this works: LLMs excel at one-off custom solutions. Official tools make sense for repeated, standardized workflows that lots of developers actually use.

๐Ÿ› ๏ธ Available Tools (All Working)

๐Ÿ“ค ExportToLLMButton Patterns

Static feeds:context="static" staticPath="exports/spec"
Current page:context="current"
Mini clipboard:mini clipboard
Developer mode:showCurlCommand showDirectUrl

โšก Implementation Reality

Minimal valid feed (2 minutes):

๐ŸŒ Basic MCP Feed - Copy & Customize๐Ÿ” Live
Replace {{placeholders}} with your actual values
{
  "feed_type": "mcp",
  "metadata": {
    "title": "{{Your Service Name}}",
    "origin": "{{https://yoursite.com}}"
  },
  "intent": "data_processing"
}
That's it. LLMs understand this immediately - no training needed.

๐Ÿ”œ Coming Soon (Realistic Timeline)

Q3 2025:
  • CLI Tools (mcp validate, test, lint)
  • WordPress & Shopify plugins
  • GitHub Actions integration
Q4 2025:
  • VS Code & JetBrains plugins
  • SDK packages (JavaScript, Python, Go)
  • CI/CD templates
2026+:
  • Browser extension for feed detection
  • Enterprise analytics & monitoring
  • Industry-specific compliance tools

๐ŸŽฏ Why Implement Now?

  • First-mover advantage: Be ready when agents go mainstream
  • Low cost: Basic MCP feed takes 15 minutes
  • Future-proofing: Web standards take 3-5 years to mature
  • It's just JSON: No complex infrastructure needed
  • Works today: All major LLMs parse it natively

๐Ÿงช Test It Right Now

Kung Fu Test: Export our spec feed and ask any LLM "What is LLMFeed?"

NEW
โ† Try this, then paste into Claude/GPT

Ready to build with the agent web?

๐Ÿฅ‹ Start Here: Train Any LLM

Transform ChatGPT, Claude, or Gemini into MCP experts in 30 seconds. Then ask them to generate everything you need.

๐Ÿง  Get Training Prompt
Philosophy: Use LLMs for custom solutions, official tools for standardized workflows with proven traction.