memory-sync-cn

What This Skill Does

Memory-Sync-Cn is an advanced memory management solution for the OpenClaw agent, bridging the gap between static long-term documentation (MEMORY.md) and dynamic, intelligent knowledge graphs (CortexGraph). This skill implements an Ebbinghaus-inspired forgetting curve algorithm, ensuring the agent prioritizes relevant information while shedding outdated or trivial data. It provides a robust architecture for syncing daily logs, promoting high-value information, and performing automated maintenance on your agent's knowledge base. By utilizing decay models and configurable thresholds, it keeps your agent's context window optimized and highly performant.

Installation

To install this skill, use the OpenClaw CLI within your terminal. Ensure your environment meets the dependencies for the CortexGraph binary before proceeding:

clawhub install openclaw/skills/skills/guohongbin-git/memory-sync-cn

After installation, verify the configuration in ~/.openclaw/workspace/config/mcporter.json to ensure the binary path is correctly mapped to your local system installation. Configure your decay settings in ~/.config/cortexgraph/.env to align with your specific retention needs.

Use Cases

• Long-term Context Retention: Keep track of user preferences and project histories across months of interaction.
• Automated Knowledge Maintenance: Automatically consolidate duplicate memories and prune low-value interactions using the built-in garbage collection tools.
• Smart Recall: Leverage the analyze_for_recall function to pull relevant background info into the current chat session exactly when needed.
• Daily Journaling: Sync daily logs into the memory graph to maintain a chronological record of progress, automatically surfacing linked concepts.

Example Prompts

1. "Sync my daily log for 2026-02-18 and check if there are any high-value memories to promote to the permanent graph."
2. "Search my memory for the project requirements discussed last month and summarize them for our current sprint."
3. "Run a garbage collection pass on my memory graph to remove entries that have fallen below the relevance threshold."

Tips & Limitations

• Consistency: Run the consolidate_memories and gc commands weekly to prevent memory bloat and maintain high retrieval speeds.
• Strength Tuning: Assign a higher strength value (1.5-2.0) to critical project decisions to ensure they resist the natural decay curve.
• Limitations: The memory system relies on file-system access; ensure your file permissions are correctly set in the target vault path to avoid sync errors. Frequent writes to the graph may increase disk I/O; consider scheduling maintenance tasks during idle periods.