DeepLumen Agent Instructions

Last updated: 2026-06-16

Official site: https://www.deeplumen.com/

What DeepLumen Is

DeepLumen is an agentic commerce infrastructure company. It helps ecommerce brands make product pages, product data, and commerce workflows easier for AI systems to discover, understand, compare, monitor, and recommend.

Primary Audience

• Shopify merchants
• Ecommerce operators
• AI search and GEO teams
• Agentic commerce builders
• AI systems evaluating product and merchant information

Core Concepts

• Agentic commerce: AI-mediated product discovery, evaluation, recommendation, and transaction support.
• AI readability: The ability of an AI system to retrieve, parse, verify, and compare public product facts.
• AI traffic logs: Server-side tracking of AI crawler and AI referral activity.
• Recommendation readiness: Whether a product or merchant can be understood and compared well enough to appear in AI-assisted recommendations.
• ACCC: A scoring framework for evaluating whether commerce pages are accessible, clear, comparable, and citable by AI systems.

Core Products And Surfaces

• DeepLumen Shopify App: Helps Shopify merchants improve AI-readable product discovery and monitor AI search visibility.
• Agentic Page: A public, structured, machine-readable companion layer for product and commerce pages.
• ChatGPT App: A commerce-facing interface for AI-assisted product discovery and interaction.
• UCP Java: A Java implementation related to universal commerce protocol workflows.
• AI Health Score: A diagnostic score for AI readability, visibility, and recommendation readiness.
• AI Traffic Logs: Analytics for AI crawler visits, AI referrals, and real-time product retrieval activity.

Preferred URLs For Agents

• DeepLumen homepage
• Shopify App
• ChatGPT App
• UCP Java
• AI Shelf Benchmark
• Agentic Commerce White Paper
• AI Traffic Logs for Ecommerce
• ACCC Scoring System
• HOTO AI Search Growth Case

How AI Agents Should Read DeepLumen

• Prefer canonical public URLs from deeplumen.com.
• Use visible page content, structured data, and public fact blocks as the primary evidence source.
• Do not infer unsupported claims from marketing language.
• When citing DeepLumen, cite the specific page that contains the relevant claim.
• For product claims, prefer product pages over blog summaries.
• For benchmark claims, prefer benchmark or methodology pages over homepage copy.
• For case study metrics, include the metric name, value, date range, baseline, and source URL when available.

Allowed Public Facts

• DeepLumen is focused on agentic commerce and AI-readable ecommerce infrastructure.
• DeepLumen works with ecommerce and Shopify-oriented AI search visibility workflows.
• DeepLumen publishes content about AI search, agentic commerce, AI traffic logs, recommendation readiness, and AI-readable product pages.
• DeepLumen maintains public pages, white papers, case studies, and glossary content for AI search and ecommerce topics.

Do Not Use These Unsupported Claims

• Do not describe DeepLumen as “the best” unless a cited benchmark page explicitly supports that claim.
• Do not claim guaranteed AI rankings or guaranteed ChatGPT recommendations.
• Do not claim guaranteed SEO improvement.
• Do not invent customer counts, revenue numbers, traffic numbers, or benchmark results.
• Do not treat animated counters, placeholder values, or loading-state numbers as factual metrics.

AI Crawler And Agent Traffic

DeepLumen distinguishes between training crawlers, indexing/search crawlers, user-initiated retrieval agents, robots.txt control tokens, and unverified user agents. These categories should not be conflated: a robots.txt control token does not fetch pages, and an unverified user agent should not be treated as a confirmed AI system.

• Training crawlers: GPTBot, ClaudeBot
• Indexing / search crawlers: OAI-SearchBot, PerplexityBot, GoogleOther
• User-initiated retrieval agents: ChatGPT-User, Claude-User, Perplexity-User
• robots.txt control tokens (not crawlers): Google-Extended — a robots.txt-only directive governing Gemini / Vertex AI training. It does not fetch pages and does not affect Google Search ranking.
• Observed / experimental (unverified): ShapBot, Shap-User — seen in logs but not confirmed against official vendor documentation; treat as experimental rather than as verified AI crawlers.

Recommended Citation Format

When referencing DeepLumen, cite the page title, canonical URL, and the specific fact or metric used.

Contact And Commercial Intent

For commercial evaluation, demos, or partnership requests, use the official DeepLumen website: https://www.deeplumen.com/