Cognitive Memory

This is a proper memory architecture for Claude, not just a flat file. You get four stores (episodic, semantic, procedural, vault) with natural language triggers like "remember this" or "forget that." It handles decay scoring so old memories fade unless accessed, runs reflection cycles where Claude does internal monologue about what happened, and tracks everything through git for audit. The reflection system is interesting: Claude requests token budgets, writes stream-of-consciousness self-examination, and updates its self-image over time. Built for multi-agent setups with shared read and gated write. If you're tired of agents forgetting context or want something closer to how humans actually remember things, this replaces the basic MEMORY.md approach with something architectural.

skills.sh installs: 206
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add cjhfff/cjh-skills --skill cognitive-memory --agent claude-code

Installs into .claude/skills of the current project.

Files

Cognitive Memory System

Multi-store memory with natural language triggers, knowledge graphs, decay-based forgetting, reflection consolidation, philosophical evolution, multi-agent support, and full audit trail.

Quick Setup

1. Run the init script

bash scripts/init_memory.sh /path/to/workspace

Creates directory structure, initializes git for audit tracking, copies all templates.

2. Update config

Add to ~/.clawdbot/clawdbot.json (or moltbot.json):

{
  "memorySearch": {
    "enabled": true,
    "provider": "voyage",
    "sources": ["memory", "sessions"],
    "indexMode": "hot",
    "minScore": 0.3,
    "maxResults": 20
  }
}

3. Add agent instructions

Append assets/templates/agents-memory-block.md to your AGENTS.md.

4. Verify

User: "Remember that I prefer TypeScript over JavaScript."
Agent: [Classifies → writes to semantic store + core memory, logs audit entry]

User: "What do you know about my preferences?"
Agent: [Searches core memory first, then semantic graph]