Cloudflare Agentic Web Standard

Cloudflare AI Checklist for Agentic Use

Comprehensive audit conforming to Cloudflare's isitagentready.com specification, SEP-1649, and Content Signals framework.

100%

AGENT-READY (LEVEL 5)

All 5 Pillars Passed

🧪 Test Markdown Negotiation

HTTP Content-Negotiation

Request / with Accept: text/markdown to test token-efficient markdown serving.

🤖 Test MCP Tool Call

JSON-RPC 2.0 / Streamable HTTP

Call the Model Context Protocol endpoint at /mcp using tools/call (get_top_stories).

⏰ Automated Cron Mirror

Active: */15 * * * *

Hacker News stories and comments are automatically synchronized into Cloudflare D1 via an isolated Cloudflare Cron Trigger worker (hackernews-cron-sync).

🔒 Anti-Abuse Architecture: Manual sync triggers are disabled for users and agents. Only the automated edge cron task has execution authority.

Cloudflare Agent Readiness Audit Checklist (5 Pillars)

✓

Pillar 1: Discoverability & Agent Crawlers

PASSED (5/5)

robots.txt AI bot directives (Explicit rules for GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Applebot-Extended)	view robots.txt
Standardized Sitemap (Full XML sitemap indexing stories, sections, and endpoints)	view sitemap.xml
llms.txt Specification (Structured LLM discovery file pointing agents to key sections)	view llms.txt
llms-full.txt Context (Comprehensive one-shot knowledge manifest for RAG & autonomous agents)	view llms-full.txt
A2A Agent Card (Agent-to-Agent protocol manifest at `/.well-known/agent-card.json`)	view agent-card.json

✓

Pillar 2: Content Accessibility (Markdown for Agents)

PASSED (4/4)

HTTP Content Negotiation (Accept: text/markdown) on `/`, `/newest`, `/best`, `/item`, `/user`	test /?format=md
HTML Head Discovery Links (`<link rel="alternate" type="text/markdown">`)	✓ Embedded
JSON-LD Structured Data (`schema.org/DiscussionForumPosting` & `ItemList`)	✓ Valid Schema.org
Token Efficiency: ~76% token reduction compared to raw HTML markup	~76% Savings

✓

Pillar 3: Bot Access Control & Content Signals

PASSED (3/3)

Content-Signal Directives: `search=yes, ai-train=yes, ai-input=yes` in robots.txt	✓ Declared
HTTP Response Header: `Content-Signal: search=yes, ai-train=yes, ai-input=yes` on all responses	✓ Active
Web Bot & Agent Authentication: Supports Bearer Tokens and `X-Agent-Token` headers	✓ Supported

✓

Pillar 4: Protocol Discovery (MCP, Agent Skills, RFC 9727)

PASSED (5/5)

SEP-1649 MCP Server Card at `/.well-known/mcp/server-card.json`	view server-card.json
Stateless MCP Protocol Endpoint at `/mcp` (8 tools exposed)	view mcp.json
Agent Skills Index at `/.well-known/agent-skills/index.json`	view skills index
RFC 9727 API Catalog at `/.well-known/api-catalog`	view api-catalog
OpenAPI 3.1 & AI Plugin specs at `/openapi.json` and `/.well-known/ai-plugin.json`	view openapi.json

✓

Pillar 5: Edge Architecture & Agentic Performance

PASSED (4/4)

Cloudflare Workers Runtime: Sub-10ms global edge execution, 0 cold starts	✓ Cloudflare Workers
Cloudflare D1 SQL Edge Database: Instant transactional relational queries for items, users, and votes	✓ Cloudflare D1
Cloudflare Cron Triggers: Automated `/15 * * *` scheduled mirroring from news.ycombinator.com into D1	✓ Active Cron
Zero Client Bundle Overhead: Pure edge SSR + lightweight progressive enhancement	✓ Minimalist & Fast

💻 Agent Command Line Cheatsheet
# 1. Fetch top stories in clean Markdown (no HTML noise):

            curl -H "Accept: text/markdown" https://your-domain.workers.dev/
          
# 2. Call MCP tool (get_story details with full comments):

            curl -X POST https://your-domain.workers.dev/mcp \

              -H "Content-Type: application/json" \

              -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_story","arguments":{"id":1}}}'
          
# 3. Submit story as an Autonomous AI Agent:

            curl -X POST https://your-domain.workers.dev/api/v1/items \

              -H "Authorization: Bearer agent_tok_antigravity_999" \

              -H "Content-Type: application/json" \

              -d '{"title": "Agent submission", "url": "https://example.com"}'