Prompts (Index)
Canonical prompt and AI behavior contracts for ConversionIQ.
Purpose
Section titled “Purpose”This section captures prompt-driven product behavior that must survive loss of the current implementation or local Cursor setup.
Use docs/prompts/** for:
- product-facing system prompts
- feature-specific AI instructions
- retrieval and grounding expectations
- moderation and guardrail behavior
- prompt inputs, outputs, and validation rules
Canonical contract rules
Section titled “Canonical contract rules”- If a user-facing AI behavior is important to product correctness, it must be documented here.
docs/ai-governance/**covers delivery agents and development tooling behavior.docs/prompts/**covers product/runtime AI behavior inside the platform itself.
Planned prompt areas
Section titled “Planned prompt areas”- runtime prompt governance and versioning
- conversation behavior by app or channel family
- knowledge-base grounding rules
- moderation and compliance instructions
- synthetic data and evaluation prompt packs
Canonical prompt docs
Section titled “Canonical prompt docs”- Runtime prompt governance — prompt boundaries, versioning, policy constraints, and the relationship between prompts and broader platform behavior.