Aaa Workflow

AAA Workflow Skill

> Orchestrator - The complete "new client" journey through the AAA Framework

Purpose

When a user says "I have a new client" or "run the AAA framework", this skill guides them through the complete Authority Acceleration journey.

When to Use

• User says "new client", "start AAA", "run AAA", "begin AAA"
• User asks "where do I start with a new client?"
• User wants the FULL workflow overview
• User seems lost in which skill to use next

The AAA Framework Overview

AAA = Authority Acceleration Agency

The framework has 4 phases:

STEP 0: CAPTURE (Deep Brand Intake)
    ↓
STEP 1: ANALYZE (Brand Foundation)
    ↓
STEP 2: ARCHITECT (Topics + Authority Positioning)
    ↓
STEP 3: ACTIVATE (Content Generation)

---

Complete Workflow

STEP 0: CAPTURE (Deep Brand Intake)

Skill: /deep-brand-intake Time: 30-60 minutes Purpose: Collect raw content samples BEFORE any analysis

Process:

1. Identify client's content sources:

• Kit/ConvertKit emails? → kit_email_extractor.py
• Podcast? → podcast_transcript_extractor.py
• YouTube? → youtube_transcript_extractor.py
• Blog? → Firecrawl scraping
• Social media? → Export/manual collection

1. Extract content to /clients/[name]/raw-content/

1. Verify minimum viable dataset:

• 10+ content pieces
• 10,000+ words
• 2+ content types

1. (Optional) Run Gemini Voice Synthesizer for large datasets:

``bash python tools/extraction/gemini_voice_synthesizer.py \ --input-dir /clients/[name]/raw-content \ --brand "[Client Name]" ``

Output:

• /clients/[name]/raw-content/ - All collected content
• /clients/[name]/source-inventory.md - What was collected

Next: Move to STEP 1 when you have enough samples

---

STEP 1: ANALYZE (Brand Foundation)

Skills: /voice-dna, /disgust-mapper, /brand-discovery Time: 45-90 minutes total Purpose: Create complete brand profile with voice patterns and boundaries

Process:

1A. Voice DNA Analysis (20-30 min)

• Skill: /voice-dna
• Input: 10-20 content samples from STEP 0
• Output: 5-layer Voice DNA Profile
• Layer 1: Structural (sentence patterns)
• Layer 2: Linguistic (word choices, phrases)
• Layer 3: Tonal (emotional register)
• Layer 4: Narrative (story patterns)
• Layer 5: Ideological (beliefs, values)

1B. Disgust Mapper (10-15 min)

• Skill: /disgust-mapper
• Input: Quick mode (3 questions) or Full mode (10 questions)
• Output: Disgust boundaries with severity scores
• What they would NEVER say
• Topics to avoid
• Tones that don't fit

1C. Brand Discovery (30-45 min)

• Skill: /brand-discovery
• Input: 39-question interactive interview OR extract from existing content
• Output: Complete Brand Profile Document
• Positioning (who, what, how, why)
• Audience deep dive
• Content strategy foundations
• Transformation journey

Output:

• /clients/[name]/voice-dna-profile.md
• /clients/[name]/disgust-profile.md
• /clients/[name]/brand-profile.md

Next: Move to STEP 2 with complete foundation

---

STEP 2: ARCHITECT (Topics + Authority Positioning)

Skills: /generate-topics, /authority-engine, /competitor-* Time: 30-60 minutes Purpose: Create strategic topic library and competitive positioning

Process:

2A. Authority Engine (15-20 min)

• Skill: /authority-engine
• Input: Brand Profile from STEP 1
• Process:
• Auto-extract niche from brand profile
• STP Analysis (Segment → Target → Position)
• Discover competitors via Perplexity
• Generate perceptual map
• Output: Competitor list + positioning strategy

2B. Competitor Analysis (20-30 min, OPTIONAL) Run if you want to learn from competitors:

• /competitor-hooks - Extract viral opening patterns
• /competitor-structures - Identify content formats that work
• /competitor-topics - Find high-engagement topic angles
• /competitor-gaps - Discover white space opportunities

2C. Topic Generation (10-15 min)

• Skill: /generate-topics
• Input: Brand Profile + Authority Engine output
• Output: 50-100 strategically positioned topics
• Scored by impact + uniqueness + urgency
• Mapped to ERISE buckets
• Organized by theme

Output:

• /clients/[name]/authority/competitors.csv
• /clients/[name]/authority/positioning-map.md
• /clients/[name]/topic-library.md

Next: Move to STEP 3 to generate content

---

STEP 3: ACTIVATE (Content Generation)

Skills: /topic-to-matrix, /trust-* Time: Ongoing (per topic) Purpose: Generate authentic, multi-platform content

Process:

3A. Topic to Matrix (5-10 min per topic)

• Skill: /topic-to-matrix
• Input: ONE topic from library + Brand Profile + Voice Meta
• Output: 1 topic → 19 pieces of content
• 1 primary asset (long-form)
• 3 Twitter threads
• 5 single tweets
• 2 LinkedIn posts
• 3 Instagram posts
• 1 Instagram carousel
• 1 YouTube Short / TikTok
• 2 email follow-ups

3B. Trust Signals (OPTIONAL - when data exists)

• Skill: /trust-signal-generator
• Purpose: Build E-E-A-T authority content
• ONLY USE IF CLIENT HAS DATA:
• /trust-case-studies - IF real client results exist
• /trust-frameworks - IF named methodology exists
• /trust-contrarian - IF defensible unique POV exists
• /trust-metrics - IF real data/numbers exist
• /trust-deep-dives - IF deep expertise to expand
• /trust-provenance - IF process to show behind-scenes

Output:

• /clients/[name]/content/[topic-slug]/ - Generated content
• Content ready for editing and publishing

---

Quick Reference Card

NEW CLIENT → CAPTURE → ANALYZE → ARCHITECT → ACTIVATE

STEP 0: CAPTURE
└── /deep-brand-intake (30-60 min)
    └── Output: raw-content/

STEP 1: ANALYZE
├── /voice-dna (20-30 min)
├── /disgust-mapper (10-15 min)
└── /brand-discovery (30-45 min)
    └── Output: voice + disgust + brand profiles

STEP 2: ARCHITECT
├── /authority-engine (15-20 min)
├── /competitor-* (optional)
└── /generate-topics (10-15 min)
    └── Output: topic library + positioning

STEP 3: ACTIVATE
├── /topic-to-matrix (per topic)
└── /trust-* (when data exists)
    └── Output: multi-platform content

---

Decision Tree

"I have a new client [NAME]"
    │
    ├── Do you have content samples?
    │   ├── NO → Start with /deep-brand-intake
    │   └── YES → How many?
    │       ├── < 10 samples → /deep-brand-intake to collect more
    │       └── 10+ samples → Move to /voice-dna
    │
    ├── Do you have a brand profile?
    │   ├── NO → Run /brand-discovery
    │   └── YES → Move to /generate-topics
    │
    └── Do you need topics?
        ├── NO (have topics) → /topic-to-matrix
        └── YES → /generate-topics first

---

The Math

Starting with ONE brand discovery session:

• 1 interview → 1 brand profile
• 1 brand profile → 100 topics
• 1 topic → 19 pieces of content
• 100 topics × 19 pieces = 1,900 pieces of content

At 3 topics/week = 12+ months of content from one session.

---

Resources

• resources/workflow-steps.md - Detailed step-by-step guide

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: zach-lloyd-dev
• Source: zach-lloyd-dev/aaa-authority-acceleration
• License: MIT
• Homepage: https://www.youtube.com/@BlackSheepSystems

Install and usage instructions live in the source repository linked above.