memos-memory-guide

What This Skill Does

The MemOS Local Memory guide is a core utility for the OpenClaw AI agent ecosystem, providing a structured interface to interact with long-term conversation history. It bridges the gap between ephemeral chat sessions and persistent knowledge storage. By utilizing a vector-based search, the skill enables agents to recall user preferences, past decisions, and specific details across multiple interactions. It manages both private memory spaces and a collaborative public layer, allowing for sophisticated multi-agent coordination. The skill is designed to augment the agent's natural "automatic recall" by providing granular tools for targeted lookups, original source retrieval, and shared information broadcasting.

Installation

To integrate this memory management capability into your OpenClaw environment, execute the following command in your terminal or command-line interface: clawhub install openclaw/skills/skills/binyuli/memos-memory-guide

Use Cases

• Personalization: Retrieving historical user context to ensure the assistant remembers names, past projects, or preferred communication styles.
• Knowledge Management: Using memory_write_public to store team decisions or project workflows so that all agents in a collaborative suite can access the same data.
• Deep Research: When a user asks a complex question about a previous interaction that wasn't covered by automatic recall, the agent can perform a surgical search using specific parameters to filter by role or keywords.
• Verification: Using memory_get to retrieve the full context of a memory chunk when a summary or search excerpt is too brief to make an informed decision.

Example Prompts

1. "What was the conclusion we reached during our meeting last Tuesday regarding the project timeline?"
2. "Search through my history for all mentions of 'Python environment setup' and summarize the steps I settled on."
3. "Save this list of architectural patterns to the public memory so other agents can reference our tech stack standard."

Tips & Limitations

• Leverage Automatic Recall: Do not overuse manual memory_search calls. The system handles standard queries effectively. Use the manual tools only when the automatic hook fails or the query is ambiguous.
• Data Sensitivity: Remember that memory_write_public exposes content to all agents in your environment. Never write sensitive API keys, passwords, or PII (Personally Identifiable Information) to public memory. Use this tool strictly for operational knowledge and team-based context.