Pollen AI by Elytron Labs

AI assistants like ChatGPT, Claude, and Perplexity are becoming a real sales
channel, but they can only recommend what they can understand. Most product
listings give them nothing to work with.

Pollen fixes that:

• AI enrichment: every product gets structured facts, an agent-optimized
  description, and search terms, generated by multimodal AI.
• Agent-readiness score: see how understandable your catalog is to
  agents, and exactly what’s missing.
• Enriched Schema.org JSON-LD: this plugin prints Pollen’s enriched
  product schema on your product pages, the data AI crawlers actually parse.
• llms.txt: served automatically from your enriched catalog, guiding AI
  crawlers and advertising your store’s agent API.
• Agent API (MCP): your catalog becomes directly queryable by AI
  assistants via the Model Context Protocol.
• AI visit analytics: see which AI crawlers visit your store and which
  ones fetch your pages to answer real shopper questions, even behind
  full-page caches.

Connecting takes one click: the plugin sends you to your own WordPress admin
to approve Pollen (WooCommerce’s built-in app authorization). No API keys to
copy.

External Services

This plugin connects to Pollen (https://pollen.elytron.in), a service
operated by Elytron Labs. It is required for the plugin’s core features and
is used in these ways:

1. Enriched catalog data (read): product pages and the llms.txt endpoint
   fetch your store’s enriched structured data and readiness score from
   Pollen, keyed by your store’s domain name.
2. AI visit analytics (write): when a known AI crawler or an AI-assistant
   referral visits your storefront, the plugin sends the visiting bot’s name,
   the page path, and your store’s domain to Pollen so your dashboard can
   show AI traffic. No visitor personal data, cookies, IP addresses, or
   human traffic details are ever sent. This reporting can be switched off
   under Settings -> Pollen.
3. Store connection: connecting uses WooCommerce’s built-in app
   authorization screen; the REST API keys WooCommerce issues are delivered
   directly to Pollen and are used to read and enrich product data only.

Data is sent only while the plugin is active. Service terms:
https://pollen.elytron.in/terms – Privacy policy:
https://pollen.elytron.in/privacy

1. Install and activate the plugin.
2. Go to Settings -> Pollen and click “Connect to Pollen”.
3. Approve the connection on the screen WooCommerce shows you. Done: Pollen
   syncs, scores, and enriches your catalog automatically.

1. 

   Home: your store on every AI channel - Google, Meta, ChatGPT, Claude, Perplexity - with live per-channel status.

2. 

   Catalog strength: five weighted score dimensions with exactly which products are missing what, worst first.

3. 

   Deep Audit: real AI assistants asked real shopper questions about your store - answers shown verbatim, cited or not.

4. 

   Proof it works: AI visits on your store, what Pollen did recently, and the one thing that needs you.

1.7.0

• Cart handoff for AI agents: a single link (?pollen_cart=ID:QTY,ID:QTY) can
  now load a multi-item cart and take the shopper straight to checkout.
  Same trust model as WooCommerce’s own ?add-to-cart URLs; stock and prices
  are enforced by your normal checkout.

1.6.0

• WebMCP for shoppers: add a free Chrome origin-trial token under Settings -> Pollen (or in your Pollen dashboard) to switch WebMCP on for your visitors today, or leave it to activate automatically when Chrome ships WebMCP.
• The WebMCP bridge is now bundled in the plugin (no external script load).

1.5.2

• AI visit analytics gains an off-switch under Settings -> Pollen.
• Settings sanitization hardening, uninstall cleanup of all plugin options
  and transients, and a WooCommerce plugin dependency declaration.

1.5.1

• Discovery links: rel=alternate links to llms.txt and the MCP server card
  in the storefront head so AI assistants can find your agent surfaces.

1.5.0

• Cache adapter: registers AI-crawler exclusions with LiteSpeed Cache,
  WP Rocket, W3 Total Cache, and WP Super Cache so AI visits are counted
  even behind full-page caches; human visitors keep full cache speed.

1.4.0

• MCP discovery surfaces: llms.txt advertises the store’s agent API;
  MCP server card served at /.well-known/mcp-server-card.
• Canonical AI-bot taxonomy shared with the Pollen dashboard.

1.3.0

• AI visit analytics split into crawl (indexing) vs retrieve (an assistant
  fetching your page to answer a live question) across all documented AI
  user agents.

1.2.0

• AI-crawler analytics: storefront visits by known AI bots are reported to
  your Pollen dashboard.

1.1.0

• Purge cached product pages when Pollen writes back enriched data, so
  crawlers always see fresh structured data.

1.0.0

• Initial release: enriched JSON-LD rendering, llms.txt serving, one-click
  connect.