No. SleekAI is configured to refuse therapy, coping advice, and diagnosis. It exists to answer practice questions and book consultations, nothing more. The boundary holds even when the visitor explicitly asks the bot to help them work through a panic attack or a difficult conversation, because that work has to come from a licensed clinician who knows the situation. The bot redirects calmly and offers a consult booking instead.

Your system prompt instructs the bot to surface 988, the Crisis Text Line, and your local emergency line when distress signals appear, and not engage with self-help content. The bot is built around your written rules - it doesn't improvise. Most practices add language that prioritises safety first, then offers to help schedule an intake, then asks if there is anything practice-related the visitor needs to know. The order matters and the system prompt enforces it.

Yes. The 28+ display conditions let you show the chatbot only on contact, fees, and clinician pages while hiding it everywhere else. For psychology practices specifically, most firms hide the bot on the blog (because reading articles often is the help-seeking activity itself) and on the about page (where the visitor is researching, not booking). Showing it only on intake-relevant pages keeps the experience appropriate.

Yes, in your WordPress database. You decide retention through the plugin settings, typically 7 to 30 days for mental health practices given the sensitivity. Conversations stay on your server, and OpenAI or Anthropic calls use your API key directly without passing through any SleekAI proxy. Most practices forward conversation summaries to their EMR through a webhook and purge raw logs on a strict schedule appropriate for sensitive material.

Yes. SleekAI is multibot, so each clinician's bio page can have a chat instance scoped to their modalities, populations served, and availability. This is particularly useful in group practices where the partners practise materially different approaches: a CBT-only clinician's bot does not need to discuss EMDR or IFS, and an EMDR-trained trauma specialist's bot needs different intake language than a general anxiety practitioner.

Absolutely. The system prompt is plain text in the WordPress admin. You can review, version, and edit it anytime, and you can set up a review cadence (most practices audit quarterly). Because the prompt is the safety floor for all clinical boundaries, the audit is a meaningful exercise rather than a checkbox. Most practices also keep a redline history of prompt edits so they can review what changed and why if a conversation later becomes a question.

Yes. Tag each clinician's availability and modality preference for telehealth versus in-person in custom fields, and the bot matches accordingly. Some clinicians do trauma work only in person, some do CBT only by telehealth, and the bot respects those preferences. The visitor's stated preference (telehealth, in-person, either) flows into the matching logic and into the booking handoff.

It can quote the plans you accept and explain superbills and out-of-network reimbursement, all from your insurance page. It does not verify a specific patient's benefits, since verification requires a call to the carrier with the patient's member ID, and the result depends on plan-specific deductibles and copays. The bot makes the limit clear: it can quote what you accept in general, but verification happens at intake.