openclaw-workspace

Summary

This workspace is a personal backup and management environment for an OpenClaw AI agent named Ryuu. It includes curated skills, long-term and daily memory, and configuration files. The setup is designed for use with WSL2, supports Telegram integration, and leverages models via OpenRouter. The structure is focused on continuity, privacy, and tailored automation to assist its primary user, Robert.

Included Assets

• Agent identity and philosophy:
  • SOUL.md — Defines the AI agent’s personality, boundaries, and continuity principles.
  • AGENTS.md — Workspace orientation and session startup procedures.
• Memory:
  • MEMORY.md — Curated, long-term memory of key learnings and user information.
  • memory/ — Dated daily interaction logs and session data (2026-03-09.md, etc.), plus session state files.
• Skills:
  • skills/ — Installed skill definitions, CLI setup docs, and config files for:
    • Todoist — Task management.
    • blogwatcher — Blog & RSS monitoring.
    • github — GitHub CLI integration.
• Workspace config:
  • TOOLS.md — Local environment notes (device names, Telegram IDs, hostnames, etc.).

How to Use

1. Workspace Orientation
   • Start each session by reviewing SOUL.md to align with agent principles and voice.
   • Follow the process in AGENTS.md for context gathering:
     • Read SOUL.md, (optionally USER.md), recent daily memory files, and in main sessions, MEMORY.md.
2. Skills
   • Refer to each skill’s SKILL.md for installation, setup, and usage (Todoist, Blogwatcher, GitHub).
   • Ensure required CLI tools are available in your shell environment.
3. Memory Handling
   • Use the memory/ directory for raw, daily logs.
   • Update MEMORY.md in main sessions for curated insights and decision records.
   • Do not expose MEMORY.md content outside of trusted, private sessions.
4. Personalization & Notes
   • Use and update TOOLS.md for hardware IDs, environment details, Telegram IDs (such as Robert's), etc.
   • Maintain privacy and update boundaries as described in SOUL.md.

Notes

• The workspace is intended for individual, private use and should be handled securely, especially regarding session and memory files.
• Only load sensitive files like MEMORY.md in main, trusted sessions to avoid data leaks.
• Skills are kept modular; local information and secrets should remain in TOOLS.md, separate from shared skill definitions.
• For any external actions (sending messages, emails, etc.), always confirm and respect user boundaries as laid out in SOUL.md.