Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.memanto.ai/llms.txt

Use this file to discover all available pages before exploring further.

Memory Types Reference

Complete reference for all 13 semantic memory types in Memanto.

Why Memory Types?

Memory types serve two purposes:
  1. Organization - Group related memories instead of throwing everything into a generic “notes” bucket.
  2. Filtering - Query specific types (e.g., retrieving only “commitments” or “preferences”).

Quick Reference

TypeUse CaseExample
factObjective information”User is in Finance dept”
preferenceLikes/dislikes”Prefers email contact”
decisionChoices made”Chose PostgreSQL”
commitmentPromises/obligations”Will deliver Friday”
goalObjectives”Reach 10K users”
eventThings that happened”Meeting at 2pm”
instructionRules/procedures”Validate input always”
relationshipConnections”Alice manages Bob”
contextSituational info”In Q1 planning”
learningLessons learned”Users need help”
observationPatterns noticed”Traffic peaks Fridays”
errorMistakes to avoid”Skip validation=bug”
artifactDocuments/files”Q3 budget.xlsx”

Detailed Definitions

fact

Objective, verifiable information about the world or domain.
  • Static information
  • Not opinions or preferences
  • Verifiable claims
  • Core knowledge
Examples:
  • “Paris is the capital of France”
  • “Database uses PostgreSQL”
  • “Team has 5 engineers”
CLI:
memanto remember "Paris is the capital of France" --type fact

preference

User or system likes, dislikes, or preferences.
  • Opinions about preferences
  • User/system choices
  • Style choices
  • Communication preferences
Examples:
  • “Prefers dark mode”
  • “Email over phone”
  • “Concise responses”
CLI:
memanto remember "Prefers email contact" --type preference

decision

Important choices made that affect future behavior.
  • Technology choices
  • Process decisions
  • Strategic choices
  • Why decisions
Examples:
  • “Chose TypeScript for codebase”
  • “Decided to use AWS”
  • “Selected Agile methodology”
CLI:
memanto remember "Chose PostgreSQL for database" --type decision

commitment

Promises or obligations made.
  • Deliverables
  • Promises to keep
  • Obligations
  • Accountability
Examples:
  • “Will deliver report by Friday”
  • “Committed to 10% growth”
  • “Promised 24/7 support”
CLI:
memanto remember "Will deliver API docs by Friday" --type commitment

goal

Objectives to achieve.
  • Future targets
  • KPIs
  • Milestones
  • Aspirations
Examples:
  • “Reach 1M users”
  • “Improve performance by 50%”
  • “Launch MVP by Q2”
CLI:
memanto remember "Launch MVP by March 31" --type goal

event

Something that happened or will happen.
  • Historical events
  • Meetings
  • Incidents
  • Milestones
Examples:
  • “Had meeting with CEO”
  • “System outage on March 20”
  • “Project launched successfully”
CLI:
memanto remember "Team meeting at 2pm" --type event

instruction

Rules, guidelines, or procedures to follow.
  • Policies
  • Best practices
  • Procedures
  • Constraints
Examples:
  • “Always validate input”
  • “Follow REST conventions”
  • “Require code review”
CLI:
memanto remember "Always validate user input" --type instruction

relationship

Connections or relationships between entities.
  • People relationships
  • System dependencies
  • Business relationships
  • Hierarchies
Examples:
  • “Alice manages Bob”
  • “Project X depends on API Y”
  • “Customer referred by John”
CLI:
memanto remember "Alice manages Bob" --type relationship

context

Contextual information about the current situation.
  • Seasonal context
  • Business phases
  • System states
  • Environmental factors
Examples:
  • “In Q1 planning phase”
  • “Budget season is active”
  • “System in maintenance”
CLI:
memanto remember "We're in Q1 planning" --type context

learning

Lessons learned from experience.
  • Insights
  • Lessons
  • Patterns recognized
  • Knowledge gained
Examples:
  • “Users need better onboarding”
  • “Mobile-first is essential”
  • “Documentation must be clear”
CLI:
memanto remember "Users need simpler onboarding" --type learning

observation

Something noticed or perceived.
  • Patterns observed
  • Trends noticed
  • Data observations
  • Field observations
Examples:
  • “Traffic peaks on Fridays”
  • “Users skip step 2”
  • “Support tickets spike after releases”
CLI:
memanto remember "Traffic peaks on Fridays" --type observation

error

Mistakes to avoid in the future.
  • Warnings
  • Lessons from failures
  • Anti-patterns
  • Known issues
Examples:
  • “Avoid deprecated API”
  • “Skip validation=bug”
  • “Don’t hardcode secrets”
CLI:
memanto remember "Never skip input validation" --type error

artifact

Important documents, files, or references.
  • Documents
  • Code files
  • Resources
  • References
Examples:
  • “Q3 budget spreadsheet”
  • “API documentation file”
  • “Customer contract”
CLI:
memanto remember "Q3 budget spreadsheet" --type artifact

Selection Guide

Does it describe something that happened?
  → event

Is it a choice or decision?
  → decision

Is it a promise or obligation?
  → commitment

Is it a rule or guideline?
  → instruction

Is it someone's preference?
  → preference

Is it an objective fact?
  → fact

Is it a future goal?
  → goal

Is it a lesson learned?
  → learning

Is it a pattern observed?
  → observation

Is it a warning or mistake?
  → error

Is it a connection between things?
  → relationship

Is it situational info?
  → context

Otherwise, ask: "Is it a document/file?"
  → artifact

Best Practices

DO

  • Use specific types for better organization
  • Choose the most specific type available
  • Be consistent in your selections
  • Filter queries by type when appropriate

DON’T

  • Overuse generic “fact” type
  • Store everything as one type
  • Mix unrelated information in one memory
  • Ignore type distinctions
  • Use wrong type for convenience
  • Skip commitment type for promises
  • Forget about error type for lessons

Filtering by Type

CLI

# Get only commitments
memanto recall "what must we do" --type commitment

# Get only preferences
memanto recall "what does user like" --type preference

# Get only decisions
memanto recall "why did we choose this" --type decision

API

response = httpx.get(
    f"{base}/agents/{agent}/recall",
    params={
        "query": "What must we do?",
        "memory_type": "commitment"
    },
    headers=headers
)