openclaw-workspace

Summary

A structured workspace for building, extending, and managing AI assistant agents in OpenClaw using Java. This workspace provides core documentation, agent setup guides, persistent memory systems, example multi-agent integrations, and modular skills for web, document, and workflow automation.

Included Assets

• AGENTS.md
  Agent onboarding and workspace usage instructions.

• MEMORY.md
  Long-term memory template and usage guide (personal context, preferences tracking).

• SOUL.md
  Defines assistant philosophy, boundaries, and best practices.

• TOOLS.md
  Local environment and tool configuration notes.

• luoshen-demo/README.md
  Example: Multi-agent system demo using AgentScope-Java.

• skills/tencent-docs/doc/doc_format/README.md
  Document formatting module for text-to-XML workflows.

• docs/

  • meeting-2026-03-27-agent-sdk.md: Agent SDK/architecture meeting notes.
  • meeting-2026-03-27-full-transcript.md: Full transcript of a development meeting.

• memory/
  Daily logs and persistent memory files (date-stamped journal, heartbeats).

• skills/
  Modular skill definitions and integration guides. Includes skills for agent browser automation and skill discovery/install flows (agent-browser, find-skills).

How to Use

1. Read the onboarding docs
   Start with AGENTS.md for session and memory management flow. Check SOUL.md for expected assistant behavior and code of conduct.
2. Memory management
   • Track important info and curated long-term memory in MEMORY.md.
   • Log daily context and reminders in memory/YYYY-MM-DD.md.
   • Use heartbeat-state.json for health/status checks.
3. Customize workspace tools
   Add local environment details (devices, SSH, TTS prefs) in TOOLS.md.
4. Skills and integrations
   Explore modular skills in skills/ (see SKILL.md in each skill folder for usage and install details).
5. Review and build on example demos
   See luoshen-demo/README.md for a multi-agent Java application example (based on AgentScope-Java).
6. Consult documentation
   Meeting notes and technical decisions are detailed in docs/.

Notes

• Workspace is optimized for clarity and persistent agent memory.
• Modularity is encouraged: skills and personal tools are decoupled for easy extension.
• Java is the primary development language.
• Memory files often include personal and sensitive context—handle accordingly.
• This repository is intended as a starting point; extend or adjust structure as needed for your assistant agents and skillsets.