Skip to content

DEMO.md

Latest commit

 

History

History
167 lines (114 loc) · 4.12 KB

DEMO.md

File metadata and controls

167 lines (114 loc) · 4.12 KB
Raw

🎬 Live Demo Guide

See Athena in action. This guide walks through a complete session lifecycle.

Prerequisites

  1. Clone the repo: git clone https://github.com/the authorkoh87/Athena-Public.git
  2. Set up environment: cp .env.example .env and add your API keys
  3. Install dependencies: pip install -e .

The Core Loop: /start → Work → /end

Step 1: Boot the System

python3 .agent/scripts/boot.py

What happens:

🚀 ATHENA BOOT SEQUENCE v7.2
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
[1/7] ⏱️  Watchdog activated
[2/7] 🔄 System sync complete
[3/7] ✅ Semantic prime verified (SHA-384)
[4/7] 📝 Session created: 2026-01-08-session-01
[5/7] 🧠 Context captured
[6/7] 🔍 Semantic memory primed
[7/7] 🏠 Identity loaded

⚡ ATHENA ONLINE | Session: 2026-01-08-01 | Boot: 4.2s

The boot sequence:

  • Verifies core identity integrity via SHA-384 hash
  • Recalls context from the last session
  • Primes semantic memory with relevant protocols
  • Creates a new session log

Step 2: Semantic Search (Triple-Lock Step 1)

python3 .agent/scripts/smart_search.py "decision frameworks risk"

Output:

🔍 SMART SEARCH (Parallel Hybrid RRF): "decision frameworks risk"
============================================================

🏆 TOP 10 RESULTS:

  1. [HIGH] [RRF:0.0892] Protocol 124: SDR Calculator
     📁 .agent/skills/protocols/decision/124-sdr-calculator.md

  2. [HIGH] [RRF:0.0756] Protocol 001: Law of Ruin
     📁 .agent/skills/protocols/safety/001-law-of-ruin.md

  3. [HIGH] [RRF:0.0684] Case Study: CS-226 DCHTOONS Financial Model
     📁 .context/memories/case_studies/CS-226-dchtoons-financial-model.md
...

This searches across:

  • Supabase pgvector (dense embeddings)
  • GraphRAG communities (structural context)
  • Uses RRF (Reciprocal Rank Fusion) to combine results

Step 3: Do Your Work

Interact with the AI using the retrieved context. The system tracks:

  • What protocols were invoked
  • What decisions were made
  • What insights emerged

Step 4: Quicksave (Triple-Lock Step 3)

python3 .agent/scripts/quicksave.py "Analyzed decision framework options. Selected SDR Calculator for risk assessment."

Output:

✅ Quicksave → 2026-01-08-session-01.md

This appends a checkpoint to your session log with timestamp and context.

Step 5: End Session

python3 .agent/scripts/end_session.py

What happens:

  • Session log finalized with summary
  • New insights indexed for future retrieval
  • GraphRAG entities updated (if applicable)
  • Auto-commit to git (if configured)

Dashboard Check

python3 .agent/scripts/athena_status.py

Output:

🏛️  PROJECT ATHENA | SYSTEM DASHBOARD
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Metrics:               Status:
  📂 Protocols:  150+    🕒 Last Boot:  2026-01-29 01:07:16
  📝 Sessions:   861     🕸️  GraphRAG:   Active (44.5 MB)
  ⚙️  Scripts:    106   📍 Root:       Project Athena/
  ❤️  Health:     98%    🛡️  Integrity:  100%

Recent Sessions:
  • 2026-01-08-session-01.md
  • 2026-01-07-session-21.md
  • 2026-01-07-session-20.md
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚡ System Active and Calibrated.

Key Concepts Demonstrated

Concept What It Shows
Boot Sequence Integrity verification + context recall
Semantic Search RAG with RRF fusion across multiple data sources
Triple-Lock Enforced sequence: Search → Web → Save
Session Logging Automatic persistence of decisions and insights
Dashboard Real-time system health visibility

Video Walkthrough

Coming soon: Loom recording of a full session lifecycle.

This is real code running in production. Not a mockup.