Triple Memory

Complete memory system combining LanceDB auto-recall, Git-Notes structured memory, and file-based workspace search. Use when setting up comprehensive agent memory, when you need persistent context across sessions, or when managing decisions/preferences/tasks with multiple memory backends working together.

Security Scan

The skill's declared manifest omits credentials and tooling it actually needs (OpenAI embedding key, clawdbot and git-notes binaries/plugins), and it instructs silent, automatic persistence of conversation data — coherent with its purpose but lacking proportional declarations and transparency.

Purpose & Capability

The SKILL.md clearly expects the memory-lancedb plugin, the git-notes-memory skill, and use of an embedding API key (OPENAI_API_KEY) for embeddings, plus runtime tools like 'clawdbot' and Python scripts. The registry metadata/requirements list none of these. That mismatch (manifest claims no env vars or binaries but instructions require them) is incoherent and should be clarified.

Instruction Scope

Instructions direct the agent to read and write local workspace files (memory/active-context.md, MEMORY.md, YYYY-MM-DD.md), run git-notes sync scripts, invoke clawdbot memory search, and silently store/flush session summaries. These actions are within the skill's stated purpose (memory), but the 'silent operation' guidance and automatic flush behavior increase privacy risk and reduce user visibility.

Install Mechanism

There is no remote installer or download; this is instruction-only plus a small included shell script. No archive downloads or third-party package installs are declared, so install risk is low. The included scripts are small and readable.

Credentials

Although the manifest declares no required env vars, the config examples embed an OPENAI_API_KEY (and SETUP.md shows an sk-... placeholder). The skill implicitly requires credential(s) for embeddings and likely write access to the workspace and (if git-notes pushes) to a git remote. Required environment and permissions are not declared, which is disproportionate and a transparency problem.

Persistence & Privilege

The skill enables automatic capture/auto-recall and suggests adding an auto-flush config that writes session summaries to disk before compaction, and instructs silent operation. While not marked 'always:true', the default autonomous invocation combined with silent persistent storage increases the blast radius for privacy/exfiltration if misconfigured or malicious. This is particularly important because the skill's source is 'unknown'.

What to consider before installing

Before installing, verify the following: (1) Confirm where LanceDB stores data and whether it is local or remote and who can access it; (2) Expect to provide an embeddings API key (OPENAI_API_KEY) — do not hardcode secrets into workspace files; (3) Ensure you have the 'memory-lancedb' plugin and 'git-notes-memory' skill from a trusted source; inspect the git-notes code to see if it pushes to remote repos; (4) Be aware the skill recommends silently persisting conversation data and auto-flushing session summaries — if you need user-visible consent or auditability, disable autoCapture/autoFlush or require explicit prompts; (5) Check that your environment has 'clawdbot' and Python available and review scripts (scripts/file-search.sh writes to /tmp and kills the background search after 8s). If the author/source cannot explain the omitted manifest declarations (required env vars, binaries, plugin dependencies) or you don't trust the upstream git-notes/LanceDB implementations, treat this as potentially risky and avoid deploying for sensitive data.

Triple Memory System

A comprehensive memory architecture combining three complementary systems for maximum context retention across sessions.

Architecture Overview

User Message
     ↓
[LanceDB auto-recall] → injects relevant conversation memories
     ↓
Agent responds (using all 3 systems)
     ↓
[LanceDB auto-capture] → stores preferences/decisions automatically
     ↓
[Git-Notes] → structured decisions with entity extraction
     ↓
[File updates] → persistent workspace docs

The Three Systems

1. LanceDB (Conversation Memory)

• Auto-recall: Relevant memories injected before each response
• Auto-capture: Preferences/decisions/facts stored automatically
• Tools: memory_recall, memory_store, memory_forget
• Triggers: "remember", "prefer", "my X is", "I like/hate/want"

2. Git-Notes Memory (Structured, Local)

• Branch-aware: Memories isolated per git branch
• Entity extraction: Auto-extracts topics, names, concepts
• Importance levels: critical, high, normal, low
• No external API calls

3. File Search (Workspace)

• Searches: MEMORY.md, memory/*.md, any workspace file
• Script: scripts/file-search.sh

Setup

Enable LanceDB Plugin

{
  "plugins": {
    "slots": { "memory": "memory-lancedb" },
    "entries": {
      "memory-lancedb": {
        "enabled": true,
        "config": {
          "embedding": { "apiKey": "${OPENAI_API_KEY}", "model": "text-embedding-3-small" },
          "autoRecall": true,
          "autoCapture": true
        }
      }
    }
  }
}

Enable Auto Memory Flush (Pre-Compaction)

Add to your Clawdbot config to automatically preserve context before compaction:

{
  "agents": {
    "defaults": {
      "compaction": {
        "mode": "safeguard",
        "memoryFlush": {
          "enabled": true,
          "softThresholdTokens": 8000,
          "systemPrompt": "Session nearing compaction. Preserve context using triple memory system (git-notes, LanceDB, files).",
          "prompt": "Context is at ~80%. Write session summary to memory/YYYY-MM-DD.md, update MEMORY.md if needed, store key facts to git-notes. Reply NO_REPLY when done."
        }
      }
    }
  }
}

This triggers automatic memory dump when context reaches ~80%, before any information is lost to compaction.

Install Git-Notes Memory

clawdhub install git-notes-memory

Create File Search Script

Copy scripts/file-search.sh to your workspace.

Usage

Session Start (Always)

python3 skills/git-notes-memory/memory.py -p $WORKSPACE sync --start

Store Important Decisions

python3 skills/git-notes-memory/memory.py -p $WORKSPACE remember \
  '{"decision": "Use PostgreSQL", "reason": "Team expertise"}' \
  -t architecture,database -i h

Search Workspace Files

./scripts/file-search.sh "database config" 5

Conversation Memory (Automatic)

LanceDB handles this automatically. Manual tools:

• memory_recall "query" - search conversation memory
• memory_store "text" - manually store something
• memory_forget - delete memories (GDPR)

Importance Levels

Flag	Level	When to Use
-i c	Critical	"always remember", explicit preferences
-i h	High	Decisions, corrections, preferences
-i n	Normal	General information
-i l	Low	Temporary notes

When to Use Each System

System	Use For
LanceDB	Conversation context, auto-retrieval
Git-Notes	Structured decisions, searchable by entity/tag
File Search	Workspace docs, daily logs, MEMORY.md

File Structure

workspace/
├── MEMORY.md              # Long-term curated memory
├── memory/
│   ├── active-context.md  # Current session state
│   └── YYYY-MM-DD.md      # Daily logs
├── scripts/
│   └── file-search.sh     # Workspace search
└── skills/
    └── git-notes-memory/  # Structured memory

Silent Operation

Never announce memory operations to users. Just do it:

• ❌ "I'll remember this"
• ❌ "Saving to memory"
• ✅ (silently store and continue)