RESEARCH LABS

WebMCP Protocols & Briefs

Technical compliance briefs, standards documentation, and reference implementations for engineering teams building machine-readable layouts.

STANDARD: WEBMCP_DEC_1.0

1. Declarative WebMCP Attributes

Conventional search crawlers index static textual pages, but autonomous AI agents (like Claude and OpenAI models) need to click buttons, fill forms, and complete actions. Declarative WebMCP attributes enable developers to decorate native HTML <form> inputs with structured descriptions.

Example code

<form 
  toolname="submit_quote_request"
  tooldescription="Submit custom development project requirements to WebMCP"
  toolaction="/api/leads"
>
  <input type="text" name="domain" required placeholder="https://example.com" />
  <input type="email" name="contact" required placeholder="email@domain.com" />
</form>

STANDARD: WEBMCP_DISC_1.0

2. /.well-known/webmcp Manifests

Placing a schema manifest file at `/.well-known/webmcp` lets search agents discover valid API parameters without scraping the raw HTML body. Think of it as a `robots.txt` specifically for generative AI actions.

JSON Payload

{
  "webmcp_version": "1.0",
  "tools": [
    {
      "name": "generate_audit_score",
      "description": "Calculates the WebMCP readiness score for a domain URL",
      "parameters": {
        "type": "object",
        "properties": {
          "url": { "type": "string", "format": "uri" }
        },
        "required": ["url"]
      }
    }
  ]
}

STANDARD: LLMS_TXT_0.8

3. llms.txt Context Indices

The `llms.txt` file format is a newly emerging markdown standard designed to serve concise, structured information specifically to LLMs during their lookup phase. We include complete `llms.txt` maps with every professional integration.

Markdown formatting

# WebMCP Agency Index
High-level structured markdown overview designed for ingestion by LLM models.

## Core Products
* WebMCP Audit: Full 35-point compatibility scanner.
* Custom Implementation: Integration of declarative headers.

## API Endpoints
* POST /api/scan: Triggers raw compliance analysis (JSON body).

Technical Library

WHITE PAPERAI Agent Transaction Loops

How autonomous models navigate forms and finalize purchases without human assistance.

CRAWLER INDEXAI Bot User-Agent Catalog

Reference map of active scrapers (GPTBot, ClaudeBot, PerplexityBot) and recommended robots.txt templates.

CHROME DEV RULESChrome v146 Prompt API

Exposing local scripts directly to the browser-built-in on-device Gemini models.

Subscribe to Protocol Updates

Join 1,200+ technical developers receiving bi-weekly code alerts regarding upcoming AI crawler regulations.