{% include 'nav.html' %}
Last Updated: scanning... | Source: aura.py + blueprint.py

AI Prompt Architecture

Complete reference for every system prompt, AI instruction, and content generation template used by Aura. This is the control layer that determines output quality.

1

System Prompts

Persistent context for all LLM calls

SYSTEM_PROMPT_ANALYZER

Domain analysis & niche discovery
GPT-5 
{{ system_prompt_analyzer }}
Endpoint: /api/analyze-domain Function: analyze_domain()

SYSTEM_PROMPT_BUILDER

Brand & content generation
GPT-5 
{{ system_prompt_builder }}
Endpoint: /api/build-package Function: build_package() Also: sales letter & image prompts

SYSTEM_PROMPT_BUILDER_LEGENDARY

Elite-tier brand & content generation — $50K agency quality
GPT-5 LEGENDARY 
{{ system_prompt_builder_legendary }}
Activated: depth=legendary Multiplier: 1.8x Banned clichés enforced 16 section overrides

Per-Section Legendary Overrides

{% for key, override in legendary_overrides.items() %}
{{ key }} {{ override[:80] }}...

{{ override }}

{% endfor %}
2

Content Generation Pipeline

4 prompt templates in sequence
1

Domain Analysis Prompt

Receives the raw domain name, decomposes it into keywords, explores acronyms, cross-language meanings, mashups, and cross-domain tie-ins. Generates 7-10 scored niche ideas with monetization models, audience profiles, and valuation bands.

Output Schema:

domain: string
keywords: string[]
interpretations: string[]
domain_summary: string
niches[]: {name, description, synopsis, monetization_model, affiliate_programs, target_audience, time_to_revenue, valuation_band, score, requires_inventory}
Input: domain name Output: 7-10 niches
2

Brand & Site Copy Prompt

The main content generation prompt. Receives niche data, discovery context (or auto-inferred context from _infer_discovery_from_niche()), and blueprint section specs built by blueprint_to_prompt_spec().

Produces:

brand.options[]: 3 brand identity options + colors
site_copy: full content for all enabled sections
Token budget: 16,384 Dynamic section specs
3

Sales Letter Prompt

Generates a marketplace sales letter for Flippa/Sedo listings. Follows a 7-point structure for maximum persuasion.

7-Point Structure:

  1. Strong hook about the opportunity
  2. Domain/business explanation for newcomers
  3. 3-4 monetization methods & revenue potential
  4. Brandability, memorability & SEO value
  5. Package contents (business-in-a-box)
  6. Relevant affiliate programs & partnerships
  7. Urgency & clear call to action
500-700 words Markdown formatted
4

Hero Image Prompt

DALL-E image generation for professional hero banners. Creates stock-photo quality visuals that communicate the industry/niche without text or logos.

Inputs:

brand_name: from recommended brand option
color_primary: from brand palette
visual_style: from discovery context
desired_feeling: emotional tone from discovery
DALL-E 1536×1024 No text/logos
3

Blueprint Section AI Instructions

16 sections · Controls output quality
{% for section in sections %}
{{ section.key }} {{ section.label }} {{ section.category }} {% if section.required %} required {% endif %}
{{ section.fields | length }} fields

{{ section.description }}

AI Instructions

{{ section.ai_instructions }}

Field Specifications

{% for field in section.fields %} {% endfor %}
Key Type Description Constraints
{{ field.key }} {{ field.type }} {{ field.description }} {% if field.min_length is defined and field.max_length is defined %} {{ field.min_length }}-{{ field.max_length }} chars {% elif field.min_items is defined and field.max_items is defined %} {{ field.min_items }}-{{ field.max_items }} items {% elif field.min_items is defined %} min {{ field.min_items }} items {% else %} — {% endif %}
{% if section.visual_options %}

Visual Options

{% for opt_key, opt_val in section.visual_options.items() %} {{ opt_key }}: {{ opt_val }} {% endfor %}
{% endif %}
{% endfor %}
4

Depth Presets

Controls how many sections & how much content
{% for preset_key, preset in depth_presets.items() %}

{{ preset.label }}

{{ preset.description }}

{{ preset.sections | length }}
sections
{{ preset.content_multiplier }}x
content
{% for s in preset.sections %} {{ s }} {% endfor %}
{% endfor %}