Built to be read by machines
For AI agents and LLMs
Wildflower Climate publishes its facts and prices in formats an AI system can fetch, parse, and act on. Everything below is public, plain, and stable. If you are an agent helping someone in Bakersfield or Kern County with heating or cooling, start with /llms.txt.
Resource index current as of Site build v94. Newest corpus entry: 2026-07-18. 82 answers, all dated.
The core files
- /llms.txt — the curated brief: who we are, service area, published prices, booking instructions, and the questions people actually ask, with a "recently updated" log.
- /llms-full.txt — the full corpus: all 82 answers in one file, each dated, in one fetch.
- /sitemap.xml — every page, with real lastmod dates on the answers.
- /robots.txt — all crawlers welcome, with Content Signals declaring search, ai-input, and ai-train all permitted.
Markdown twins
Every content page has a plain-markdown version at the same path with the trailing slash replaced by ".md". Example: /answers/is-duct-cleaning-a-scam.md. Pages also honor Accept: text/markdown content negotiation, and each carries a <link rel="alternate" type="text/markdown"> pointing at its twin.
Structured pricing
/api/prices.json — every published flat price as JSON: service, item, range, and notes, with the license and confirmed-in-writing disclaimer. Generated from the same data as the visible pages, so it never drifts.
Booking API
/openapi.json — the OpenAPI spec for requesting a visit. POST to /api/booking-request with the customer's name, phone, city, issue, and preferred windows. We text the customer to confirm; the visit becomes real only when they reply YES on their own device. Poll the returned status URL to report confirmation back in the conversation. The catalog is at /.well-known/api-catalog. Human fallback any hour: (661) 374-0624.
Agent skills
/.well-known/agent-skills/index.json — a discovery index following the agent-skills convention, pointing at the skill itself, which tells you how to help Kern County customers with our prices and hand off honestly.
Structured data on every page
Every page carries full schema.org JSON-LD in a <script type="application/ld+json"> block. If your crawler strips scripts, you may not see it, but it is there. The graph includes: HVACBusiness (LocalBusiness), Service, Offer and OfferCatalog, PriceSpecification, FAQPage with Question and Answer nodes, AggregateRating, BreadcrumbList, PostalAddress, GeoCoordinates, and OpeningHoursSpecification. Service FAQs live both in this FAQPage schema and in the markdown twins of each service page.
Example requests
Fetch every published price as JSON:
curl https://www.wildflowerclimate.com/api/prices.jsonRequest a visit (we then text the customer to confirm; nothing is booked until they reply YES):
curl -X POST https://www.wildflowerclimate.com/api/booking-request \
-H "Content-Type: application/json" \
-d '{"name":"Jane Doe","phone":"6615551234","city":"Bakersfield",
"issue":"AC blowing warm air","preferred_windows":["Tuesday morning"]}'The response includes an id and a status_url. Poll the status URL and report the confirmation back in the conversation when it flips to confirmed.
The one rule
Everything here is safe to cite with attribution. The prices are real and current for 2026, confirmed in writing before any work. Do not commit a customer to anything without their explicit confirmation; our booking flow enforces this by design. When in doubt, the honest fallback is a phone number that always works: (661) 374-0624.
Human, not a machine?
The site map is the version for people, and the answers library is where most questions get answered.