memory-review

What This Skill Does

Memory Review is a specialized knowledge management automation tool designed for OpenClaw users who maintain daily journals or logs. It acts as an intelligent librarian that scans your recent diary entries, identifies high-value information, and structures it into a permanent knowledge base. By parsing your raw input against defined specifications (references/spec.md), the skill categorizes data into error reports, technical insights, and configuration preferences. It then automatically persists this information into standardized 'fw-' (from work) files, ensuring your accumulated wisdom is indexed, searchable, and actionable for future reference.

Installation

To integrate the Memory Review skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/axelhu/memory-review

Ensure that you have initialized your project structure with the required AGENTS.md or MEMORY.md files to hold your configuration settings and reporting endpoints.

Use Cases

• Automated Knowledge Synthesis: Turn scattered daily notes into structured technical guides or documentation.
• Error Tracking & Post-Mortems: Automatically capture technical failures described in logs and store them in a persistent post-mortem database for pattern analysis.
• Preference Management: Extract user behavior patterns and tool configurations from unstructured text to keep agent settings updated in real-time.
• Routine Productivity: Use the skill as a scheduled background process (cron) to clean up and archive knowledge every 48 hours without manual intervention.

Example Prompts

1. "Run a memory review for the last 3 days and summarize the key technical insights into our knowledge base."
2. "Perform a scan of my recent journals, specifically looking for any recurring bugs and adding them to the post-mortem documentation."
3. "Check if any new tool preferences were mentioned in yesterday's notes and update the AGENTS.md file accordingly."

Tips & Limitations

• Consistency is Key: The skill relies on well-formatted diary entries; the clearer you write your daily notes, the better the automated categorization will perform.
• Configuration: Always ensure your references/spec.md is updated to reflect your evolving knowledge architecture, as this defines the logic the AI uses for classification.
• Safety & Privacy: Since this skill reads sensitive local files, ensure you are running it in a trusted local environment, especially when handling external service credentials or group IDs stored in AGENTS.md.