A specialized technical skill for managing, auditing, and rebuilding the MemoryLab long-term memory sidecar and its associated context delivery systems.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-system-sidecar
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install memory-system-sidecar using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Memory System Sidecar?
The Memory System Sidecar is a robust implementation for the MemoryLab environment designed to handle long-term extraction, indexing, and consolidation of agent memory. Unlike basic planning tools, this skill provides the direct operational layer for the memory sidecar, serving as the bridge that feeds retrieved data into active task states and live context packets. It ensures that the agent's working memory remains synchronized with durable storage while maintaining clear boundaries between raw session history and active runtime data.
This utility is essential for developers working within the Openclaw Skills ecosystem who need to maintain high-fidelity context for AI agents. It prioritizes the integrity of raw history and transcripts as the ultimate source of truth, offering automated scripts to refresh context, rebuild database indexes, and verify system milestones through rigorous evaluation harnesses.
Memory System Sidecar Use Cases
- Refreshing live context and sidecar data from real-time laptop sessions.
- Verifying the integrity of memory system milestones and running evaluation checks.
- Rebuilding stale memory indexes when retrieval performance drifts.
- Auditing milestone status and checking for contamination boundaries in memory artifacts.
- Debugging discrepancies between status documentation and runtime memory states.
How Memory System Sidecar Works
- The system utilizes raw session history and transcripts as the primary source of truth for all memory extractions.
- The context manager acts as a hot-path compactor, while the memory sidecar performs long-term indexing and retrieval.
- A bridge mechanism injects retrieved memory into the ACTIVE_TASK_STATE without overwriting the durable storage.
- Automated shell scripts execute end-to-end refresh paths to sync the live packet with current session activities.
- Verification scripts run unit tests and evaluation harnesses to ensure system stability and milestone adherence.
Memory System Sidecar Setup
To integrate this capability into your environment, ensure you have the necessary repository wrappers accessible. Install the skill and initialize the sidecar using the provided scripts:
# Refresh the memory system from current session history
./skills/memory-system-sidecar/scripts/refresh_memory_system.sh
# Verify system status and run unit tests
./skills/memory-system-sidecar/scripts/verify_memory_system.sh
Memory System Sidecar Data Schema & Taxonomy
The skill organizes data across specific directories to maintain clear boundaries between working and durable memory as part of the Openclaw Skills architecture:
| Location | Purpose |
|---|---|
memory/ |
Contains the LIVE_CONTEXT_PACKET.md and ACTIVE_TASK_STATE.md (Working Memory). |
memory-system/ |
Stores durable memory artifacts, including MILESTONE_STATUS.md and index/memory.db. |
history/ |
Contains the raw transcripts used as the source of truth for all indexing. |
scripts/ |
Contains context-refresh and utility wrappers for system operation. |
Memory System Sidecar Advanced Features
- Automated end-to-end refresh paths that wrap complex manual command chains.
- Integration with evaluation harnesses to detect retrieval drift and verify milestone integrity.
- Support for multi-agent context synchronization via retrieval context bundles.
- Intelligent index rebuilding for
memory.dbto resolve stale-data behavior. - Advanced debugging protocols that prioritize runtime checks over static documentation.
SKILL.md
Loading
Related Openclaw Skills
Ghibli Style Image Generator
An AI-powered image generation skill that creates high-quality Studio Ghibli inspired artwork using the Neta AI API.
Media.io AI Photo Restorer
A professional AI-powered tool for automating photo restoration and quality enhancement using the Media.io OpenAPI.
Movedone
A local HTTP API integration for the Movedone kanban app that enables AI agents to manage projects, tasks, and dependencies.
CLI-Anything
A methodology and toolkit designed to turn complex software, GUI applications, and repositories into structured command-line interfaces optimized for AI agents.
a2a-Market UCP Broadcast
A specialized protocol skill for broadcasting signed buyer intents and managing node acknowledgments within a Universal Commerce Protocol (UCP) ecosystem.
Fixture — Test Data Fixture Generator
A high-performance CLI tool for generating, managing, and exporting reproducible test data fixtures using schema-based templates.
Featured*
