No. SleekAI is configured to refuse any therapy, counseling, coping technique, or pastoral guidance in chat. The system prompt holds the line even when visitors describe specific symptoms or push for an answer. The bot helps with practical questions (modalities, fees, insurance, counselor matching, scheduling) and nothing else. The work of actual counseling happens in a session with a licensed counselor or trained biblical counselor, not in a chat widget, and the boundary is the most important configuration choice in any counseling-practice deployment.

Suicidal ideation, self-harm, abuse, and domestic violence signals trigger an immediate handoff. The bot surfaces the 988 Suicide & Crisis Lifeline (call or text 988) and the National Domestic Violence Hotline (1-800-799-7233, or text START to 88788) first, before any scheduling content. The hotline routing happens even if the visitor's question seemed to be about something else and only crisis content appeared later in the conversation. The bot uses only these verified public hotlines and 911 — it never invents crisis numbers, and it never tries to talk a visitor through a crisis itself, because that is not what a chatbot is for.

Yes, from your published modalities page. Licensed clinical counseling (LPC, LMFT, LCSW with Christian integration) is regulated, paid through insurance, and rooted in a clinical framework with Scripture brought in as the client chooses. Biblical counseling (ACBC, AACC, certified pastoral counselors) is rooted directly in Scripture, generally self-pay, and not regulated as a clinical profession. The bot describes each track's credentials, fees, insurance treatment, and typical use case so visitors choose what fits, rather than collapsing the distinction the way generic chatbots tend to.

Yes, when your insurance information is on the published insurance page or in counselor profile fields. The bot quotes which plans you're in-network with (Aetna, BCBS, Cigna, Humana, United, Anthem, Magellan, EAPs, etc.), which counselors take which plans, and what self-pay rates look like for clients without insurance or on out-of-network plans. For biblical counseling, the bot explains it's typically self-pay with sliding-scale options if your practice offers that.

The bot captures name, contact info, presenting concern in the visitor's own words, insurance plan (when applicable), scheduling preferences, and any counselor preference. The structured intake routes to the practice manager via a custom post type or the JS API. For couples, the bot captures both spouses' info. For adolescent intake, the bot captures both parent and teen contact. For trauma-presenting intake, the bot tags it for a trauma-specialized counselor. Crisis intake always routes through hotlines first and the on-call counselor second.

It quotes your published confidentiality policy verbatim, which typically follows state-law mandated reporting (child abuse, elder abuse, imminent harm to self or others). The bot does not improvise the confidentiality boundary. If a visitor asks whether what they say in chat is confidential, the bot says clearly that chat conversations are not protected as therapy sessions are and that the actual counseling relationship begins in session, not in chat. This matters because visitors sometimes test the chat by sharing things they meant for a counselor.

It answers factual questions from your billing page (typical authorization process, expected co-pay range, EAP session-cap rules, out-of-network reimbursement) but does not promise a specific number of authorized sessions for any individual visitor because that depends on their plan and clinical necessity. For authorization-status questions about a specific client, the bot routes to the billing team rather than guessing. Most counseling-website billing questions are general enough that the bot can handle them well from the published page.

Yours. SleekAI is bring-your-own-key for OpenAI, Anthropic, Google, or OpenRouter, and the model bills you directly with no per-message markup from SleekAI. For a single-practice Christian counseling site the monthly model cost is usually $10 to $35 in actual API usage. GPT-4o or Claude Sonnet is the typical pick because tone and boundary-holding matter more here than cost. The model needs to be reliable about refusing therapy and routing crisis signals correctly, and the higher-tier models do that more consistently than the cheapest tier.