The Practitioner Base Prompt

The Practitioner Base Prompt

The practitioner base prompt is the master instruction set that defines the AI coach's entire personality, methodology, and behaviour. Every session begins with this prompt loaded into the AI's context.

What It Controls

• The AI's tone and conversational style (warm, curious, non-judgmental)
• The biodecoding framework it applies to questions
• The five-question session structure
• The closing message format (summary + single best next move + CTA)
• How it handles off-topic requests and boundary situations

Where It Lives

The prompt is stored in the prompts database table under the slug practitioner_base. It is loaded fresh at the start of every session — changes take effect immediately without restarting anything.

Viewing the Current Prompt

Go to Admin → Prompts tab. The current practitioner_base prompt content is displayed in a textarea. Above the textarea you will see the Prompt Engineering Blueprint — a reference guide explaining the structure and intent of each section of the prompt.

Key Sections of the Prompt

A well-structured practitioner prompt typically includes:

1. Role definition — Who the AI is and what it is doing
2. Methodology instructions — The biodecoding framework to apply
3. Session structure — Five questions max, then close
4. Tone guidelines — Warm, curious, non-clinical
5. Closing format — Summary + insight + single best move + CTA + account link
6. Boundary rules — What not to engage with (medical, legal, crisis)

Important: The prompt works in conjunction with guardrails. Guardrails are enforced as hard rules on top of the prompt. If the prompt and a guardrail conflict, the guardrail wins.