xiucheng-memory-manager_skill

Overview

This skill is a long-term memory management system for OpenClaw agents that automatically archives conversations, curates important items, and enables semantic retrieval. It captures daily logs, organizes memories by date, and maintains a consolidated long-term memory file for high-value information. The design focuses on low-friction integration so agents can save and query histories without manual bookkeeping.

How this skill works

The skill intercepts agent conversations and writes them to daily markdown files (YYYY-MM-DD.md) in a configured memory directory. It also supports a long-term MEMORY.md file where curated or summarized memories are stored. A semantic search layer indexes the archived content and returns relevant entries for natural-language queries, enabling agents to retrieve past discussions, decisions, and context quickly.

When to use it

• Preserve agent conversations for audit, compliance, or knowledge continuity.
• Provide agents with historical context to improve responses and decision-making.
• Maintain searchable records for projects, KPIs, or recurring topics.
• Automate archiving in multi-session or long-running agent workflows.
• Support handoffs between agents or human operators with context continuity.

Best practices

• Enable auto_save for consistent archival, but keep retention policies for storage control.
• Regularly curate entries into MEMORY.md to reduce noise and improve search relevance.
• Configure memory_dir and memory_file paths to match workspace backups and access controls.
• Use concise, topic-focused messages to improve semantic search accuracy.
• Periodically re-index or rebuild search embeddings after major content imports or schema changes.

Example use cases

• An agent that records daily standup conversations and later retrieves decisions about sprint priorities.
• Customer support agents that query past tickets and conversation threads to provide personalized help.
• Project agents that search historical discussions for KPI commitments and deadline rationales.
• Audit workflows that require time-stamped evidence of agent interactions and decisions.
• Onboarding agents that surface prior project summaries from MEMORY.md to accelerate new contributors.