Your First Session

Start a session

ginko start

ginko loads context from your last session (if any) and presents a readiness summary:

Ready | Hot (10/10) | Think & Build mode
Last session: Set up project structure and API routes
Next up: e001_s01_t01 - Implement auth middleware (start)

Sprint: Auth System 20%
  Follow: ADR-001, ADR-003
Branch: feature/auth (3 uncommitted files)

Your AI coding assistant reads this output and knows exactly where you left off.

Start options

ginko start --verbose        # Full session details
ginko start --compact        # Minimal output
ginko start --team           # Include team events
ginko start --auto-progress  # Auto-advance to next sprint when complete

Work with your AI

During your session, ginko provides commands to maintain context:

Log important events

ginko log "Implemented JWT validation middleware"
ginko log -c decision "Chose bcrypt over argon2 for password hashing"
ginko log -c insight -i high "Found race condition in session cleanup"

Check status

ginko status

Recalibrate when stuck

ginko vibecheck

Opens a conversational check-in: “What are we actually trying to achieve?” Helps when you’ve drifted from the goal or feel uncertain.

Check process health

ginko health

Shows a dashboard of your session quality — logging frequency, task tracking, decision recording.

Handoff

When you’re done for now:

ginko handoff

This saves:

• Current session state and progress
• Uncommitted work context
• A handoff summary for the next session

ginko handoff "Completed auth middleware, need to add rate limiting next"
ginko handoff --verbose  # Show detailed sync info

The next time you (or a teammate) runs ginko start, the full context is restored.

The session lifecycle

ginko start → work → ginko log → work → ginko handoff
     ↑                                        |
     └────────────────────────────────────────┘

Each cycle builds on the last. Your AI partner accumulates project understanding across sessions instead of starting from zero every time.