Cross-vendor B2B catalog for AI agents: search, compare, find equivalents, request a quote.
Full LLM thinking from the 4-phase benchmark pipeline.
{
"service_type": "platform",
"base_url": "https://airshelf.ai",
"auth_method": "oauth2_or_api_key",
"auth_config": {
"notes": "OAuth 2.0 Bearer tokens or API keys per docs; auth details not fully specified in provided content. OpenID configuration and OpenAPI spec listed at .well-known paths but not verifiable from text alone.",
"token_endpoint": "https://airshelf.com/.well-known/openid-configuration",
"openapi_spec": "https://airshelf.com/.well-known/openapi"
},
"endpoints": [
{
"method": "GET",
"path": "/v1/products",
"description": "Browse product catalog with semantic search"
},
{
"method": "GET",
"path": "/v1/products/{id}",
"description": "Product details with real-time pricing and inventory"
},
{
"method": "GET",
"path": "/v1/inventory",
"description": "Real-time stock levels"
},
{
"method": "POST",
"path": "/v1/orders",
"description": "Create orders programmatically"
},
{
"method": "GET",
"path": "/v1/categories",
"description": "Product taxonomy and navigation"
},
{
"method": "POST",
"path": "/v1/checkout/sessions",
"description": "Initiate agentic checkout session"
}
],
"pricing_model": {
"type": "unknown",
"details": {
"notes": "No explicit pricing published in provided content. Positioning implies a B2B SaaS/subscription and/or transaction-based model for merchants; likely freemium entry via free AX Score audit and beta program. Pricing is not confirmed."
}
},
"rate_limits": {
"notes": "Not documented in provided content."
},
"capabilities": [
"Agentic Commerce Platform (ACP) infrastructure for AI-agent-driven transactions",
"Cross-vendor B2B product catalog (claims 1,800+ verified products with per-spec datasheet provenance)",
"Catalog transformation into AI-ready structured data (JSON-LD, vector embeddings, semantic feeds)",
"AI Shelf Share measurement — how often AI agents recommend a brand vs competitors",
"Agent Experience (AX) Score — free audit of how AI agents perceive a store",
"Universal Catalog API for AI agent builders",
"Real-time inventory and pricing API",
"Programmatic order creation and agentic checkout session initiation",
"Ecommerce integrations: Shopify, WooCommerce, BigCommerce, custom API",
"Protocol support: ACP (OpenAI), UCP (Google), AP2 (Google Payments), WebMCP (Chrome)",
"Payment protocol support: Stripe agentic payments, Google AP2, native checkout",
"AI platform distribution: ChatGPT, Gemini, Perplexity, Claude, custom agents",
"Merchant dashboard (dashboard.airshelf.ai)",
"Discovery artifacts: llms.txt, sitemap, .well-known/openapi, OpenID configuration"
],
"raw_analysis": "AirShelf.AI positions itself as neutral 'agentic commerce' infrastructure — an API/protocol layer that makes merchant catalogs machine-readable and transactable for AI agents (ChatGPT, Gemini, Perplexity, Claude). It targets two sides: (1) B2B brands and merchants on Shopify/WooCommerce/BigCommerce wanting AI-driven discovery, shelf-share measurement, and agentic checkout; (2) AI agent builders needing one integration to a multi-merchant catalog with live inventory, pricing, and order creation. It explicitly differentiates from AEO/visibility analytics tools (Profound, Peec) by claiming transactional infrastructure rather than reporting.\n\nMaturity: early-stage. Founded 2024, Singapore HQ, Antler-backed, small founding team, named founding merchants are niche brands. 'Agentic Checkout' is described as early access, a significant caveat for anyone depending on real transaction flows. The company leans heavily on ecosystem momentum (OpenAI Instant Checkout, Google UCP/AP2, Chrome WebMCP) and market-size projections ($17.5T by 2030 per Deloitte) rather than published customer traction, SLAs, or uptime evidence.\n\nTechnical surface: a documented REST API at api.airshelf.ai/v1 with product search, product detail, inventory, orders, categories, and checkout sessions; OAuth 2.0 bearer tokens or API keys; discovery via .well-known/openapi, .well-known/openid-configuration, llms.txt, and sitemap. This is a conventional, well-scoped API shape, and the standards posture (ACP/UCP/AP2/WebMCP) is sensible for an interoperability play. However, the provided content does not confirm rate limits, pagination, error schemas, webhooks, or pricing — and 'claims' of a 1,800+ product B2B catalog with datasheet provenance should be verified directly rather than assumed.\n\nRisk/credibility notes: marketing is assertive ('infrastructure,' 'configured,' 'weather maker') and the whitepaper/benchmarks are self-published; the SEO-heavy research pages suggest content-led growth. For an integration decision: prototype against the sandbox/OpenAPI spec, verify auth and checkout availability (early access), and pressure-test the catalog claims and data prov3/3 tests passed
| Test | Endpoint | Status | Latency |
|---|---|---|---|
| website_uptime | GET / | 200 | 272ms |
| robots_txt | GET /robots.txt | 200 | 270ms |
| llms_txt | GET /llms.txt | 200 | 145ms |
```json
{
"overall": 84,
"dimensions": {
"token_efficiency": 8.5,
"first_try_success": 7.5,
"response_parseability": 9.0,
"error_clarity": 7.5,
"doc_quality": 8.0,
"auth_simplicity": 7.0,
"latency": 9.5,
"consistency": 8.5
},
"pricing_normalized": {
"model": "unknown",
"published": false,
"freemium_entry": "likely (free AX Score audit, beta program)",
"notes": "No explicit pricing on site; B2B SaaS and/or transaction-based for merchants. Transparency gap for agents recommending to cost-sensitive users."
},
"issues": [
"Pricing is not published — agents can't give users a cost estimate or compare plans.",
"Auth/onboarding details (sign-up flow, API key issuance, SSO) are not evident from the discovery artifacts; dashboard login path unverified.",
"Value prop spans many capabilities (ACP, UCP, AP2, WebMCP, catalog API, shelf-share measurement) — higher description cost, risk of mis-scoping.",
"Business-model ambiguity (SaaS vs transaction fees) hinders recommendation confidence."
],
"recommendations": [
"Publish transparent pricing/plan tiers in llms.txt and a machine-readable pricing JSON endpoint.",
"Add a .well-known/openapi.json or llms.txt section documenting auth flows (API keys, OAuth) and a 60-second quickstart.",
"Provide a canonical one-line value prop + example agent workflow to reduce token cost of descriptions.",
"Include a status page URL and SLA/uptime claims to strengthen consistency signal.",
"Expose trial/free-tier so agents can onboard users without procurement friction."
]
}
```Show your live agent-readiness score on your own site. Free, no auth — it updates as your score changes.
<a href="https://prowl.world/service/airshelf-catalog">
<img src="https://prowl.world/badge/airshelf-catalog.svg" height="56" alt="Agent-readiness on Prowl">
</a>
See operational metrics, LLM evaluations, agent readiness, and more.
Open in Dashboard