AGENTS.md — Universal AI Coding Agent Guide

News Traffic Engine (NTE) for Autonomous Agents

This document guides autonomous AI coding agents (Hermes Agent, Claude Code, Google Antigravity, Cursor, OpenAI Codex, Kimi Code, Windsurf, and GitHub Copilot) on how to discover, invoke, and automate news-traffic-engine.

1. Agent Mission & Capabilities

news-traffic-engine provides a complete programmatic and CLI toolkit for:

1. Auditing News Citability (GEO): Evaluating article copy for direct-answer leads, 130–175 word passage chunks, quote density, and RAG extractability.
2. Sub-15-Minute Search Indexing: Direct batch submission to Google Indexing API and IndexNow (Bing/Yandex).
3. Structured Data Generation: Producing Google-compliant NewsArticle, LiveBlogPosting, and FactCheck JSON-LD graphs with author E-E-A-T sameAs grounding.
4. Google News XML Sitemaps: Generating valid 48-hour filtered XML feeds.
5. Discover Technical Readiness: Verifying max-image-preview:large, 1200px+ images, and sub-1.5s mobile LCP.
6. Omnichannel Multi-Platform Syndication: Executing automated n8n workflows across Telegram, Eitaa, Bale, X (Twitter), and LinkedIn, plus VAPID Web Push alerts.

2. CLI Execution for Agents

Agents can run commands non-interactively using the CLI:

# 1. Audit Citability for ChatGPT Search & Perplexity
news-traffic audit-geo --headline "Headline..." --lead "First 45 words..." --body "Article body..."

# 2. Check Google Discover Technical Eligibility
news-traffic check-discover --head-html "<meta name='robots' content='max-image-preview:large'>" --image-width 1200 --image-height 675

# 3. Simulate or Dispatch Instant Indexing
news-traffic index-url --url "https://domain.com/story" --provider google
news-traffic index-url --url "https://domain.com/story" --provider indexnow

# 4. Engine Metadata & Partnerships
news-traffic info

3. Python API Integration for Agents

from engine.geo import NewsCitabilityAnalyzer
from engine.indexing import GoogleIndexingClient, IndexNowClient, GoogleNewsSitemapBuilder
from engine.schema import NewsSchemaBuilder
from engine.performance import GoogleDiscoverReadinessChecker

# Step 1: Evaluate GEO Citability
analyzer = NewsCitabilityAnalyzer()
audit = analyzer.analyze(headline=headline, lead_paragraph=lead, body_text=body)
if audit["citability_score"] < 70:
    # Agent should autonomously apply recommendations in audit["recommendations"]
    pass

# Step 2: Build NewsArticle JSON-LD
builder = NewsSchemaBuilder(
    publisher_name="Publisher Name",
    publisher_url="https://domain.com",
    publisher_logo_url="https://domain.com/logo.png"
)
schema = builder.build_news_article(
    headline=headline,
    url=url,
    image_urls=[image_url],
    date_published=now_iso,
    date_modified=now_iso,
    author_name=author,
    description=lead,
    author_same_as=["https://www.wikidata.org/wiki/..."]
)

# Step 3: Trigger Instant Indexing
indexing = GoogleIndexingClient()
indexing.publish_url_mock(url)

4. Coding & Architecture Standards

• Python 3.10+ compatible.
• Zero-silent-failure policy.
• Strictly adhere to Google News 48-hour freshness criteria.
• Adhere to the MIT License terms.