Skip to main content

​
Memory Commands

Store, retrieve, and manage memories using the MEMANTO CLI.

​
memanto remember

Store a single memory. 
memanto remember TEXT [OPTIONS]
Arguments:
  • TEXT - Memory content (required)
Options:
  • -t, --type TEXT - Memory type (default: fact)
  • --title TEXT - Optional title
  • --confidence FLOAT - Confidence 0-1 (default: 1.0)
  • --batch PATH - Load memories from JSON file
Examples: Simple memory: 
memanto remember "Paris is the capital of France"
With type: 
memanto remember "Customer prefers email" --type preference
With title: 
memanto remember "Will deliver report Friday" --type commitment --title "Delivery Deadline"
Lower confidence: 
memanto remember "Thinks customer is in finance" --type fact --confidence 0.7
Batch from file: 
memanto remember --batch batch.json
Memory Types: fact, preference, decision, commitment, goal, event, instruction, relationship, context, learning, observation, error, artifact

​
memanto recall

Retrieve memories using semantic search. 
memanto recall QUERY [OPTIONS]
Arguments:
  • QUERY - Search query (required)
Options:
  • -t, --type TEXT - Filter by type
  • -l, --limit INTEGER - Max results (default: 5, max: 100)
  • --sort TEXT - Sort by: recency, confidence (default: recency)
  • --as-of DATETIME - Recall as of timestamp
  • --changed-since DATE - Changed since date
  • --current-only - Only current (non-superseded) memories
Examples: Basic recall: 
memanto recall "What about the customer?"
Filter by type: 
memanto recall "preferences" --type preference
Specific limit: 
memanto recall "customer info" --limit 10
Sort by confidence: 
memanto recall "important facts" --sort confidence
Temporal queries: 
# As of yesterday
memanto recall "status" --as-of "2025-03-29T12:00:00Z"

# Changed in last week
memanto recall "updates" --changed-since "2025-03-23"

# Only current (not archived)
memanto recall "commitments" --current-only
Output: 
Found 3 memories:

1. Customer prefers email contact
   Type: preference
   Confidence: 0.98
   Created: 2025-03-26 09:00:00 UTC

2. Will deliver report by Friday
   Type: commitment
   Confidence: 1.0
   Created: 2025-03-28 14:30:00 UTC

3. In PST timezone
   Type: fact
   Confidence: 0.95
   Created: 2025-03-25 10:15:00 UTC

​
memanto answer

Get AI-powered answers grounded in memories. 
memanto answer QUESTION [OPTIONS]
Arguments:
  • QUESTION - Question to answer (required)
Options:
  • -l, --limit INTEGER - Memories to use (default: 10)
Examples: Simple question: 
memanto answer "How should we communicate?"
With custom limit: 
memanto answer "What's the customer's profile?" --limit 20
Output: 
Based on your memories:

The customer prefers email communication and is in the PST timezone.
You should send email during business hours PST (8am-5pm).
The customer appreciates concise responses, so keep communications brief.

​
memanto daily-summary

Generate a comprehensive daily summary of all memories. 
memanto daily-summary [OPTIONS]
Options:
  • -a, --agent TEXT - Specific agent (default: active agent)
Examples: Summary for current agent: 
memanto daily-summary
Summary for specific agent: 
memanto daily-summary --agent customer-support
Output: 
═══════════════════════════════════════════════════════
Daily Memory Summary: customer-support
Date: 2025-03-30
═══════════════════════════════════════════════════════

<Icon icon="chart-simple" /> STATISTICS
├── Total Memories: 42
├── New Today: 8
├── Updated: 2
└── Archived: 1

<Icon icon="folder" /> BY TYPE
├── facts: 15
├── preferences: 8
├── events: 12
├── commitments: 4
└── decisions: 2

<Icon icon="star" /> KEY FACTS
- Customer Alice is in Finance
- Order #12345: shipped
- Payment: credit card on file

​
memanto conflicts

Detect and resolve contradictory memories. 
memanto conflicts [OPTIONS]
Options:
  • -a, --agent TEXT - Specific agent (default: active agent)
  • --threshold FLOAT - Confidence threshold (default: 0.7)
Examples: Check for conflicts: 
memanto conflicts
Stricter threshold: 
memanto conflicts --threshold 0.9
Specific agent: 
memanto conflicts --agent customer-support
Output: 
Found 1 Potential Conflict:

CONFLICT 1 (Confidence: 0.88)
─────────────────────────────────
Contradiction Type: Direct Opposite
Memory A (older):  "Customer is in finance team"
Memory B (newer):  "Customer moved to marketing team"

[Interactive resolution]
Choose action:
[1] Mark Memory A as outdated
[2] Mark Memory B as outdated
[3] Keep both
[4] View details

Choose (1-4): 1
✓ Memory A marked as superseded

​
Memory File Commands

​
memanto memory export

Export all agent memories to file. 
memanto memory export [OPTIONS]
Options:
  • --agent TEXT - Specific agent (default: active)
  • --format TEXT - Format: md, json (default: md)
  • --output PATH - Output file path
Examples: Export to markdown: 
memanto memory export
Export to JSON: 
memanto memory export --format json
Custom output: 
memanto memory export --output backup.md
Specific agent: 
memanto memory export --agent customer-support --format json
Output: Creates file with all memories: 
# Agent: customer-support
Date: 2025-03-30
Total Memories: 42

## Facts
- Customer Alice Johnson (2025-03-20)
- Works in Finance (2025-03-20)

​
memanto memory sync

Sync memories to MEMORY.md for Claude Code integration. 
memanto memory sync [OPTIONS]
Options:
  • --agent TEXT - Specific agent
Examples: Sync active agent: 
memanto memory sync
Sync specific agent: 
memanto memory sync --agent customer-support
Output: 
✓ Memories synced to MEMORY.md
  Agent: customer-support
  Memories: 42
  File: ./MEMORY.md

​
Batch Operations

​
Create batch.json

[
  {"content": "Alice Johnson", "type": "fact"},
  {"content": "Works in Finance", "type": "fact"},
  {"content": "Prefers email", "type": "preference"},
  {"content": "PST timezone", "type": "fact"},
  {"content": "Will deliver report Friday", "type": "commitment"}
]

​
Store Batch

memanto remember --batch batch.json
Output: 
✓ Batch stored successfully
  Stored: 5 memories
  Status: queued for processing

​
Workflow: Complete Memory Cycle

# 1. Create agent
memanto agent create customer-service

# 2. Activate
memanto agent activate customer-service

# 3. Store memories
memanto remember "Customer: Alice Johnson" --type fact
memanto remember "Prefers email" --type preference
memanto remember "Finance department" --type fact

# 4. Recall
memanto recall "Tell me about Alice"

# 5. Get AI answer
memanto answer "How should we communicate with Alice?"

# 6. Daily summary
memanto daily-summary

# 7. Check conflicts
memanto conflicts

# 8. Export backup
memanto memory export --format json --output backup.json

​
Tips & Best Practices

​
DO

  • Use specific memory types
  • Include titles for clarity
  • Regular exports for backup
  • Check conflicts weekly
  • Use batch import for large datasets

​
DON’T

  • Store duplicates
  • Generic types (use fact, preference, etc.)
  • Ignore conflicts
  • Forget to extend sessions
  • Store sensitive data

​
Next Steps