LaunchKit · 2026
Back to Skills

engram-memory

Persistent semantic memory for AI agents — local, fast, free.

1
1.5k downloads
by @dannydvm

Setup & Installation

openclaw skills install @dannydvm/engram-memory

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install engram-memory

What This Skill Does

Local semantic memory system for AI agents. Stores facts, decisions, preferences, and events with biological-style decay, so frequently accessed memories stay relevant. Supports relationship graphs and cross-session context without any cloud dependency.

Runs entirely locally with no API keys or cloud accounts, so memory persists instantly and privately without external service dependencies.

When to use it

  • Recalling a client's preferred tech stack mid-session
  • Storing a decision to pause a campaign for later retrieval
  • Searching past conversation history by topic or type
  • Linking two memories with a 'contradicts' or 'supersedes' relationship
  • Exporting and restoring agent memory across machines

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User asks: remember that we decided to use Postgres instead of MySQL for this project

AGENT
  1. 1Search existing memories for any prior database decisions
  2. 2Store new decision memory: 'We decided to use Postgres over MySQL'
  3. 3Tag memory with relevant labels (e.g., database, architecture)
  4. 4Run auto-relate to link it to any related tech stack facts
  5. 5Confirm storage with engram stats
OUTPUT

Decision stored, salience-boosted, and linked to related project facts for future recall