Skip to main content
Mega Brain’s session management system allows you to save your work and resume exactly where you left off. This guide covers session workflows, context preservation, and best practices.

Session Overview

Sessions preserve:
  • Current task and status
  • Pending files and processing queue
  • Recent operations and results
  • Important file references
  • User notes and context
Sessions are stored in /logs/LIVE-SESSION/ with both human-readable Markdown and structured JSON formats.

Saving Sessions

Basic Save

/save
System prompts: “What do you want me to save as context?”
User input:
"I've ingested 3 Cole Gordon videos and processed 2 of them.
Still need to process the third one and then extract DNA."
Saved session:
# SESSION: 2026-03-06 14:32

## Status
IN_PROGRESS

## Last Task
Processing Cole Gordon materials
- Completed: 2/3 videos
- Pending: 1 video (closing-masterclass-part3.txt)

## Next Steps
1. Process inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt
2. Extract DNA: /extract-dna cole-gordon
3. Verify agent enrichment

## Important Files
- inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt
- knowledge/dossiers/persons/DOSSIER-COLE-GORDON.md

## Notes
User wants to complete Cole Gordon DNA extraction before
moving to Alex Hormozi materials.

Automatic Saves

The system auto-saves context:
1

After Pipeline Completion

When /process-jarvis completes, session automatically updates with:
  • Source processed
  • Artifacts created
  • Agent enrichment status
2

After DNA Extraction

When /extract-dna completes, session records:
  • DNA quality metrics
  • Number of items per layer
  • Average weight
3

After Conclave

When /conclave completes, session saves:
  • Decision recommendation
  • Confidence score
  • Action items

Resuming Sessions

Basic Resume

/resume
═══════════════════════════════════════════════
           RESUMING SESSION
═══════════════════════════════════════════════

📅 Last Activity
Processing Cole Gordon materials
Saved: 2026-03-06 14:32 (2 hours ago)

🟡 Status
IN_PROGRESS

Completed:
✓ Processed 2/3 videos
  - closing-masterclass-part1.txt → CG004
  - closing-masterclass-part2.txt → CG005

Pending:
⚠ 1 video remaining:
  - inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt

🎯 Next Steps
1. Process remaining video:
   /process-jarvis inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt
   
2. Extract DNA after processing:
   /extract-dna cole-gordon
   
3. Verify agent enrichment:
   /agents --knowledge "Cole Gordon"

📁 Important Files
- inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt
- knowledge/dossiers/persons/DOSSIER-COLE-GORDON.md

📝 Notes
User wants to complete Cole Gordon DNA extraction
before moving to Alex Hormozi materials.

───────────────────────────────────────────────
Want to continue from where you left off? (Y/n)

Resume with Context

System loads:
  1. Task status - What was being worked on
  2. File references - Exact paths to continue
  3. Pending queue - What’s left to do
  4. User notes - Custom context saved

Session Files

File Locations

logs/LIVE-SESSION/
├── CURRENT-SESSION.md    # Human-readable session log
├── CONTEXT.json          # Structured session data
└── session-history/     # Archive of past sessions
    ├── 2026-03-05.md
    ├── 2026-03-04.md
    └── 2026-03-03.md

CONTEXT.json Structure

{
  "last_task": "Processing Cole Gordon materials",
  "status": "in_progress",
  "pending": [
    "Process closing-masterclass-part3.txt",
    "Extract DNA for Cole Gordon",
    "Verify agent enrichment"
  ],
  "next_steps": [
    "/process-jarvis inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt",
    "/extract-dna cole-gordon",
    "/agents --knowledge 'Cole Gordon'"
  ],
  "important_files": [
    "inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt",
    "knowledge/dossiers/persons/DOSSIER-COLE-GORDON.md"
  ],
  "notes": "User wants to complete Cole Gordon before Alex Hormozi",
  "saved_at": "2026-03-06T14:32:00Z",
  "statistics": {
    "files_processed": 2,
    "files_pending": 1,
    "insights_extracted": 23,
    "agents_enriched": 3
  }
}

Advanced Session Management

Check Session Status

# View current session without resuming
/log
Output:
## Current Session

**Status:** IN_PROGRESS
**Started:** 2026-03-06 14:00
**Last Updated:** 2026-03-06 14:32
**Duration:** 32 minutes

**Progress:**
- Files processed: 2
- Files pending: 1
- Insights extracted: 23
- Agents enriched: 3

**Recent Operations:**
1. [14:15] Ingested cole-gordon/closing-part1.txt
2. [14:18] Processed CG004 → 23 chunks, 12 insights
3. [14:25] Ingested cole-gordon/closing-part2.txt  
4. [14:28] Processed CG005 → 21 chunks, 11 insights
5. [14:32] Saved session

View Session History

# List past sessions
ls logs/LIVE-SESSION/session-history/

# Read specific session
cat logs/LIVE-SESSION/session-history/2026-03-05.md

System Briefing

Get complete system status:
/jarvis-briefing
═══════════════════════════════════════════════
JARVIS: Operational Briefing
═══════════════════════════════════════════════

🟢 Health Score: 92/100

📚 Knowledge Base:
   Experts:        3 (Alex Hormozi, Cole Gordon, Sam Ovens)
   Dossiers:       8 (5 persons, 3 themes)
   Playbooks:      2
   Total Insights: 127
   DNA Schemas:    3

📄 Last Processing:
   "Cole Gordon - Closing Masterclass Part 2" (CG005)
   Processed: 2026-03-06 14:28
   Result: 21 chunks, 11 insights

🤖 Active Agents: 22
   C-Level:  4 (CRO, CFO, CMO, COO)
   Sales:    8 (CLOSER, BDR, SDS, LNS, ...)
   Minds:    3 (Alex, Cole, Sam)
   Conclave: 3 (Critic, Advocate, Synthesizer)

📥 Inbox Status:
   Pending: 1 file
   └─ cole-gordon/MASTERCLASS/closing-masterclass-part3.txt

📆 Current Session:
   Status: IN_PROGRESS
   Task: Processing Cole Gordon materials
   Progress: 2/3 complete

⭐ Next Action:
   /process-jarvis inbox/cole-gordon/MASTERCLASS/closing-masterclass-part3.txt

═══════════════════════════════════════════════

Session Workflows

Multi-Day Project

1

Day 1: Ingest Materials

# Ingest multiple videos
/ingest https://youtube.com/watch?v=video1
/ingest https://youtube.com/watch?v=video2
/ingest https://youtube.com/watch?v=video3

# Save progress
/save
Input: “Ingested 3 Cole Gordon videos. Will process tomorrow.”
2

Day 2: Resume and Process

# Resume where left off
/resume

# Process files
/process-jarvis inbox/cole-gordon/video1.txt
/process-jarvis inbox/cole-gordon/video2.txt

# Save progress
/save
Input: “Processed 2/3 videos. Third one tomorrow.”
3

Day 3: Complete and Extract DNA

# Resume
/resume

# Process final video
/process-jarvis inbox/cole-gordon/video3.txt

# Extract DNA
/extract-dna cole-gordon

# Save completion
/save
Input: “Completed Cole Gordon processing and DNA extraction. Ready for next expert.”

Interrupted Work

Scenario: Pipeline running, need to stop immediately
# Stop pipeline (Ctrl+C)
^C

# Save current state
/save
Input:
"Pipeline was at Phase 4 (Insight Extraction) when stopped.
Need to rerun from beginning when resuming.
File: inbox/alex-hormozi/offer-framework.txt"
Resume later:
/resume
# System shows exactly where you stopped
# Rerun from beginning:
/process-jarvis inbox/alex-hormozi/offer-framework.txt

Session Best Practices

Save Frequently

Save after completing each major step:
  • After ingesting batch
  • After processing material
  • After DNA extraction
  • Before/after Conclave

Use Descriptive Notes

Include:
  • What you completed
  • What’s pending
  • Any blockers
  • Context for decisions

Reference File Paths

Always include:
  • Exact file paths
  • Source IDs (CG001, AH002)
  • Artifact locations

Track Dependencies

Note what needs to complete first:
  • Process before extract DNA
  • Get approval before implement
  • Verify before scaling

Troubleshooting Sessions

Issue: Resume shows wrong contextSolution:
# Check current session file
cat logs/LIVE-SESSION/CURRENT-SESSION.md

# If corrupted, restore from history
cp logs/LIVE-SESSION/session-history/2026-03-06.md logs/LIVE-SESSION/CURRENT-SESSION.md

# Resume again
/resume
Issue: Lost progress (no recent save)Recovery:
# Check JARVIS briefing for last processing
/jarvis-briefing

# Check file registry for what was processed
cat system/REGISTRY/file-registry.json

# Check inbox for pending files
/inbox

# Reconstruct context and save
/save
Issue: Multiple overlapping sessionsSolution:
# Review session history
ls -lt logs/LIVE-SESSION/session-history/

# Pick the correct session
cat logs/LIVE-SESSION/session-history/[date].md

# Copy to current
cp logs/LIVE-SESSION/session-history/[date].md logs/LIVE-SESSION/CURRENT-SESSION.md

# Resume
/resume

Viewing Processing Logs

Different log types for different purposes:
Current session:
cat logs/LIVE-SESSION/CURRENT-SESSION.md
Session history:
ls logs/LIVE-SESSION/session-history/
cat logs/LIVE-SESSION/session-history/2026-03-06.md

Next Steps

Start Your First Session

  1. Ingest materials: /ingest [source]
  2. Save session: /save
  3. Resume anytime: /resume
  4. Continue workflow

Process Pipeline

Learn the 5-phase processing workflow

Work with Agents

Query enriched agents after processing

Build docs developers (and LLMs) love