Skip to content

Latest commit

 

History

History
121 lines (91 loc) · 3.36 KB

File metadata and controls

121 lines (91 loc) · 3.36 KB

Recommended Workflow

Quick Start (Single Person Analysis)

1. Export chat from any platform → Save to data/chats/[name]/
   - WhatsApp: Chat → ⋮ → Export → Include media
   - Telegram: Desktop → ⋮ → Export chat history
   - Slack/Discord/iMessage: Use platform export tools
   - Or just drop any text chat log in the folder
2. cd ~/Desktop/chat-analyzer
3. claude (start Claude Code session)
4. /analyze-chat
5. /profile-person
6. /sales-insights
7. Ask follow-up questions

Multi-Agent Architecture

Main Agent (You interact with this)

  • Orchestrates analysis
  • Synthesizes results from sub-agents
  • Answers your questions
  • Generates final reports

Sub-Agent Types & When to Use

Agent Subagent Type Use For
Explorer Explore Finding files, cataloging data, understanding structure
Analyzer general-purpose Processing chat data, pattern detection
Researcher general-purpose Web searches for frameworks, validation
Planner Plan Designing complex multi-step analyses

Parallel Processing Pattern

For large datasets, run agents in parallel:

User: Analyze this 10,000 message chat history

Claude orchestrates:
├── Agent 1: Analyze messages Jan-Mar
├── Agent 2: Analyze messages Apr-Jun
├── Agent 3: Analyze messages Jul-Sep
└── Agent 4: Analyze messages Oct-Dec

Then synthesizes all results into unified profile

Research Enhancement Pattern

For deeper insights:

User: Create a sales strategy for this person

Claude orchestrates:
├── Agent 1: Analyze their expressed interests
├── Agent 2: WebSearch for "selling to [personality type]"
├── Agent 3: WebSearch for "[interest category] trending products"
└── Agent 4: Research objection handling techniques

Then combines into personalized strategy

Session Management

Starting Fresh (New Person)

# Archive current data
mkdir -p data/archive/$(date +%Y%m%d)-current
mv data/chats/* data/archive/$(date +%Y%m%d)-current/
mv data/media/* data/archive/$(date +%Y%m%d)-current/

# Start new Claude session
claude

# Load new data and analyze

Comparing People

1. Ensure both profiles exist in analysis/profiles/
2. /compare-sessions
3. Select profiles to compare

Example Prompts

Initial Analysis

  • "Analyze all chat data in the data folder"
  • "Create a comprehensive profile of this person"
  • "What are their main interests and pain points?"

Sales Focus

  • "What products would this person be interested in?"
  • "How should I pitch [specific product] to them?"
  • "What objections might they raise about [topic]?"
  • "When is the best time to reach out?"

Deep Dive

  • "How has their communication style changed over time?"
  • "What triggers positive responses from them?"
  • "What topics should I avoid?"
  • "How do they make decisions?"

Research Enhancement

  • "Research effective sales techniques for their personality type"
  • "Find trending products in [their interest area]"
  • "What psychological frameworks explain their behavior?"

Data Privacy Best Practices

  1. Don't sync to cloud: Keep folder local or use encrypted storage
  2. Delete after analysis: Remove raw chat exports when done
  3. Anonymize profiles: Replace names with codes for sensitive analysis
  4. Review before sharing: Profiles contain inferred personal data