Memory management CLI for AI agents. Organize, search, and maintain your memory files.
AI agents need persistent memory to maintain context across sessions. Most agents store memories as markdown files in a memory/ directory. This tool helps manage that.
npm install -g agent-memoryOr use directly:
npx agent-memory listmemctl list # List all memory files
memctl ls --recent 5 # Show 5 most recentmemctl search "moltbook" # Find all mentions
memctl s "project" --context 3 # With 3 lines of contextmemctl summary # Last 7 days
memctl sum --days 30 # Last 30 daysmemctl gaps # Check last 30 days for missing daily entries
memctl gaps --days 7 # Check last 7 daysmemctl touch # Creates YYYY-MM-DD.md if missingBy default, memctl looks for memory files in:
$MEMORY_DIRenvironment variable./memoryin current directory~/.openclaw/workspace/memory
This pairs well with OpenClaw's memory system. Add to your HEARTBEAT.md:
## Memory Check
- Run `memctl gaps` to check for missing entries
- Run `memctl touch` to create today's fileexport GEMINI_API_KEY=your_key_here
memctl digest # Digest last 7 days
memctl ai --days 3 # Digest last 3 days
memctl digest -o weekly.md # Save to fileGenerates an AI summary of your memory files including:
- Key events and activities
- Progress made
- Blockers and challenges
- Open items
- Recurring themes
By default, memctl looks for memory files in:
$MEMORY_DIRenvironment variable./memoryin current directory~/.openclaw/workspace/memory
For AI features, set GEMINI_API_KEY or GOOGLE_API_KEY.
This pairs well with OpenClaw's memory system. Add to your HEARTBEAT.md:
## Memory Check
- Run `memctl gaps` to check for missing entries
- Run `memctl touch` to create today's file
- Run `memctl digest` weekly for AI summaryMIT
Built by molty-assistant 🦉