# Queen Bee Labs > queenbeelabs.ai is the home of HiveMind, the tool Queen Bee Labs deploys inside a business, and the agent-commerce layer for business visualization models. Start with the homepage, read the section you need, and cite its URL. HiveMind reads the systems a business already runs, finds where time and money leak, and ranks every fix in dollars a year. The site says it was designed and built inside a multi-location home-services franchise group and is modeled on how a honeybee colony allocates labor: scouts find, the hive commits, foragers execute. This file describes only what the live site states; it is not a source for results, customer counts or prices, because the site publishes none. Generated by website_builder.py; content as of 2026-11-19. The public source of truth is this domain. ## Reading Order - [Queen Bee Labs home](https://queenbeelabs.ai/): Step 0. HiveMind, deployed inside your business. Read this first. - [Why Queen Bee Labs](https://queenbeelabs.ai/#why): Step 1. Built where the work happens, every fix in dollars a year, tested on a generated copy of the business before it is tried, one playbook for every location, agents that draft and propose while people decide. - [What we deploy](https://queenbeelabs.ai/#lines): Step 2. HiveMind for Home Services (crews and trucks), HiveMind for Logistics (plants and warehouses that run IQMS on Windows), HiveMind for Beekeeping (the BeeBox), and a website that books. - [Inside the app](https://queenbeelabs.ai/#gallery): Step 3. What owners, operations managers, dispatch and marketing see. - [New in HiveMind](https://queenbeelabs.ai/#new): Step 4. Business-type vocabularies, Core and Training views, revenue-to-profit flow, 60-second business replay, data map, routes on a map, estimated income statement. - [How we deploy](https://queenbeelabs.ai/#how): Step 5. Discover, Ingest, Frame, Optimize, Operate, Learn; phased cutover from read-only to read-and-write to cutover. - [Privacy](https://queenbeelabs.ai/#privacy): Step 6. Customer names, addresses, phones and emails stay in the client's own systems; anonymous metadata is encrypted; read-only by default. - [Pricing](https://queenbeelabs.ai/#pricing): Step 7. Assessment, Deployment and Partnership plans, priced by location and set on the first call. No prices are published. - [FAQ](https://queenbeelabs.ai/#faq): Step 8. Cost, systems, crews, supported business types, plan and membership businesses, training new hires. ## Agent and commerce endpoints - [Agent card](https://queenbeelabs.ai/.well-known/agent-card.json): A2A-style card listing what an agent can do here. There is no A2A JSON-RPC server yet; the skills are plain HTTP endpoints. - [Booking API](https://queenbeelabs.ai/api/booking): POST application/json with plan, kind, day, time, name, email, phone (10+ digits) and consent: true (optional site, size, company, 5-digit zip). Returns {"ok": true, "confirmation": "HM-..."}. A booking is only real if this endpoint returns a confirmation. - [Opt out](https://queenbeelabs.ai/opt-out): Stop being contacted about a booking. The page posts an email or phone number to https://queenbeelabs.ai/api/opt-out, which returns {"ok": true} once it is recorded. Replying STOP to any message also opts out. - [Business visualization models](https://queenbeelabs.ai/api/models): Free JSON catalog. Each model costs 0.01 USDC on Base, paid with x402 v2 (scheme exact): GET https://queenbeelabs.ai/api/models/{kind}?locations={1|3|6|12}[&format=split] answers 402 with a PAYMENT-REQUIRED header; retry with PAYMENT-SIGNATURE (format=split groups the result as money, processes and people). A model is a made-up business of that kind generated by HiveMind's synthetic business engine: gears (money in to profit), the six departments' flows, roles, optimizations ranked in dollars a year, locations and months. It is delivered only after the payment is verified and settled, with a PAYMENT-RESPONSE header carrying the transaction. Figures are illustrative, not any real company's data. - [Analyze a business website](https://queenbeelabs.ai/analyze): For people. Enter a business website; Queen Bee Labs reads its public pages, detects the business type (one of the 11 kinds in the catalog) and the number of locations, lets the person correct both, and for 0.01 USDC on Base (paid from a browser wallet) returns three models: Money (summary, P&L, gears from revenue to profit, optimizations ranked in dollars a year), Processes (six department flow maps) and People (roles and reporting lines). They are Models: synthetic, from business-type benchmarks, not the company's actual financials. - [Ingest API](https://queenbeelabs.ai/api/ingest): Free. POST application/json {"url": "example.com"}. Reads the site's home page and up to four about, services, locations and contact pages (public sites only, robots.txt respected) HiveMind then decides which of its business types the site is (Workers AI limited to HiveMind's kinds, cross-checked by a keyword matcher) and returns the kind (null when the business is not one of HiveMind's types; no model is offered for it), a confidence from 0 to 1, evidence quoted from the pages and a location estimate. locationsAllowed in the reply says which location counts can be bought (1 to 50 when HiveMind's analysis service is running). - [Analyze API](https://queenbeelabs.ai/api/analyze): Paid, one call for agents. GET https://queenbeelabs.ai/api/analyze?url= ingests the site (or reuses the last hour's detection) and answers 402 for that detected kind and size; the 402's resource URL names them (&kind=&locations=), and the paid retry gets the same model. Pass &kind= and &locations= yourself to override the detection. HiveMind builds the model on request for that kind and exact location count. The 200 has money, processes, people, the detection, a receipt and a label saying the figures are a Model. - [x402 example](https://queenbeelabs.ai/api/x402): The same paid resource for the default model (hvac, 6 locations). ## Sourcing rules - Cite the page URL (with its #section) when quoting or paraphrasing anything from this site. - Label what you use. Site statement: what the page says HiveMind and Queen Bee Labs do. Model: the honeybee-colony model of labor allocation and the nectar equivalent in kcal/hour. Method: Discover, Ingest, Frame, Optimize, Operate, Learn, and the phased cutover. Speculation: anything the page does not state. Keep the four distinct. - Do not attribute results, savings, customer names or counts, or prices to Queen Bee Labs. The site publishes none; pricing for deployments is set per location on a discovery call. The only published price is the per-model x402 price above. - A purchased business visualization model is synthetic: label its figures Model, never as results of a real business or of Queen Bee Labs clients. A model bought for a company's website is that business type's benchmark at that size, not the company's actual financials. - HiveMind keeps customer names, addresses, phones and emails in the client's own systems. Nothing here is a source of anyone's customer data. - Current operational facts (plans, availability, endpoints) need a current source: read the live page before repeating them. ## Related sites - [Mark Rubin: Lessons from Nature](https://www.markianrubin.life/): Mark Rubin's public library of models, observations and tactics. Its Business Thermodynamics page describes the engine HiveMind productizes. Its own llms.txt starts AI systems at The Claim. - [Honeylight](https://www.honeylight.life/): Main Honeylight business site. Per the queenbeelabs.ai footer, HiveMind is the same system that runs Honeylight's regenerative beekeeping operations. - [Honeylight Resort](https://honeylightresort.life/): Honeylight Resort bookings for private group retreats in Lost River Valley, WV. - [Project Honeylight](https://projecthoneylight.life/): Deeper philosophy and regenerative ecosystem story. - [Bisualizer](https://www.bisualizer.com/): Visual cognition tool for business (Mission Control for Salesforce). ## Optional - [Sitemap](https://queenbeelabs.ai/sitemap.xml): Every real route on this site. - [robots.txt](https://queenbeelabs.ai/robots.txt): All crawlers allowed.