The system prompt explicitly forbids diagnosis, treatment, or therapy framing and instructs the bot to refer to licensed clinicians for clinical concerns. We provide a sample scope-of-practice prompt for coaches that you can drop in as a starting point. Any question that drifts toward symptoms, medication, or trauma processing gets met with a referral plus a reminder of where coaching helps and where it doesn't.

Yes. Configure the prompt to detect self-harm, suicide, or acute crisis cues and respond with 988 in the US or your local equivalent before mentioning programs, pricing, or booking. The crisis response should be hard-coded so it never depends on the model picking the right tone in the moment, and it should run as the first check on every reply.

If your certifications, NBHWC, IIN, ACSM, or anything else, are listed on your About page or coach bios, the bot reads and cites them accurately. It also knows what each credential does and doesn't authorise, so it won't claim a health coach prescribes meal plans or treats clinical anxiety. Update one bio and the next conversation reflects the change.

Yes. Hand off bookings via deep link or webhook to Calendly, Acuity, SavvyCal, TidyCal, or any tool with a public booking URL. The bot collects the prospect's preferred slot in conversation, then either passes them to your booking link with the slot pre-selected or fires a webhook that creates the event server-side. Most coaches use deep links because they're easier to debug.

Logs stay in your WordPress database, attached to the page the conversation started on. Bring-your-own-key means the prompts go directly to OpenAI, Anthropic, or Google with no third-party processor in between. You're responsible for HIPAA-style controls on the WordPress side; the plugin doesn't add a new vendor to your data map, which is what most coaches care about most.

Yes. Every conversation is logged in WordPress with timestamp, page, and pre-call answers attached. Many coaches scan the log five minutes before the discovery call to clock the prospect's stated sleep, stress, or burnout pattern, then walk into the call already knowing the top three blockers instead of using the first 15 minutes to surface them.

Only if you tell it to. The default prompt asks about goals, timeline, and budget first, then recommends the program that matches, including group programs when 1:1 is overkill. If you want it to lead with the 12-Week Reset, set that as the recommended path in the prompt and add a fallback to the lower-tier option when budget signals appear.

It declines and routes. Medication, supplement protocols, eating-disorder treatment, trauma processing, and acute crisis all sit outside coaching scope, and the system prompt tells the bot to say so plainly and offer the appropriate referral. If a prospect insists, the bot repeats the boundary and offers to take their email so a human can follow up the next morning.