Skip to main content

​
Quick Start (5 Minutes)

Get MEMANTO working with real memory operations in under 5 minutes.

​
Prerequisites

  • MEMANTO installed (pip install memanto)
  • Moorcheh API key configured (memanto setup)
  • Terminal access

​
1. Check Status (30 seconds)

Verify everything is set up: 
memanto status
You should see: 
Moorcheh API Key:  ✓ Valid
Moorcheh Connection: ✓ Connected
Active Agent:      (none - activate one to start)
Active Session:    (none - activate one to start)

​
2. Create Your First Agent (30 seconds)

memanto agent create quick-demo
Output: 
Agent 'quick-demo' created successfully.
Agent ID: quick-demo

​
3. Activate Session (30 seconds)

memanto agent activate quick-demo
Output: 
Session activated for agent 'quick-demo'
Session Token: eyJ... (6-hour session)
Session expires: 2025-03-30 20:30:00 UTC
Verify activation: 
memanto status
You should now see your active agent and session.

​
4. Store Your First Memory (1 minute)

Store a simple fact: 
memanto remember "Paris is the capital of France" --type fact
Output: 
Memory stored successfully
Memory ID: mem_abc123xyz
Type: fact
Store a preference: 
memanto remember "I prefer responses in JSON format" --type preference
Store a commitment: 
memanto remember "Will complete the project by Friday" --type commitment

​
5. Recall Memories (1 minute)

Retrieve memories using semantic search: 
memanto recall "What is the capital of France?"
Output: 
Found 1 memory:

1. Paris is the capital of France
   Type: fact
   Confidence: 1.0
   Created: 2025-03-30 16:30:00 UTC
Recall by semantic similarity (not exact match): 
memanto recall "What format should responses be in?"
Output: 
Found 1 memory:

1. I prefer responses in JSON format
   Type: preference
   Confidence: 0.95
   Created: 2025-03-30 16:31:00 UTC

​
6. Get AI-Powered Answers (1 minute)

Generate answers grounded in your memories: 
memanto answer "Where is the capital of France and what format should I use?"
Output: 
Based on your memories:

The capital of France is Paris. I've stored that you prefer responses
in JSON format, so here's a structured response:

{
  "country": "France",
  "capital": "Paris"
}

​
7. View Daily Summary (optional)

See all your memories and any conflicts: 
memanto daily-summary
Output: 
Daily Summary for Agent 'quick-demo'

Total Memories: 3
├── facts: 1
├── preferences: 1
└── commitments: 1

Key Facts:
- Paris is the capital of France

Preferences:
- I prefer responses in JSON format

Commitments:
- Will complete the project by Friday

No conflicts detected.

​
8. Detect Contradictions (optional)

Check for conflicting memories: 
memanto conflicts
If you add a contradictory memory: 
memanto remember "Paris is in Germany" --type fact
Then check conflicts: 
memanto conflicts
Output: 
Potential Conflicts Detected

Conflict 1:
- Memory 1: "Paris is the capital of France" (fact)
- Memory 2: "Paris is in Germany" (fact)
- Confidence: 0.87 (likely conflicting)

Recommendation: Remove or update one memory

​
Next Steps

​
For CLI Users

​
For Developers (Programmatic Access)

Want to use MEMANTO via REST API or Python? → See First Agent Setup

​
Explore Features

​
Common Scenarios

# Store customer information
memanto remember "Customer name: Alice Johnson" --type fact
memanto remember "Alice works in finance" --type fact
memanto remember "Alice prefers meetings on Tuesdays" --type preference

# Recall all related memories
memanto recall "Tell me about Alice"

​
Track Decisions

# Record a decision
memanto remember "Chose PostgreSQL for new service" --type decision

# Later, recall the decision
memanto recall "Why did we pick PostgreSQL?"

​
Manage Commitments

# Store a commitment
memanto remember "Deliver API documentation by April 15" --type commitment

# Check what you've committed to
memanto recall "What did I commit to?"

​
Learn From Errors

# Record what went wrong
memanto remember "NPM install failed due to old Node version" --type error

# Recall lessons learned
memanto recall "Node version issues"

​
Session Management

Sessions expire after 6 hours by default. Extend when needed: 
# Check current session
memanto session info

# Extend session
memanto session extend --hours 6

# Or deactivate and reactivate
memanto agent deactivate
memanto agent activate quick-demo

​
Storage

Memory is stored in Moorcheh.ai (serverless, no local storage needed). To export memories for backup: 
memanto memory export
This creates a markdown file with all your agent’s memories at ~/.memanto/exports/agent_memory.md.
Done! You’ve learned: ✓ Create agents ✓ Store memories ✓ Recall semantically ✓ Get AI answers ✓ Detect conflicts ✓ Manage sessions Now explore advanced features in the guides!