openclaw-workspace

Summary

This workspace provides structured documentation and configuration guidelines for agent persona, memory management, tool integration, and local environment notes for OpenClaw setups. It enables consistent personality, memory, and operational practices for agent-based workflows.

Included Assets

• AGENTS.md — Agent session and memory operation manual
• MEMORY.md — Long-term memory rules (in Chinese)
• SOUL.md — Agent persona and core behavioral guidelines (in Chinese)
• TOOLS.md — Template for environment-specific configuration
• ppt-design/node_modules/ — Documentation and metadata for language and tooling dependencies:
  • Babel, Puppeteer, QuickJS, Node.js, and Yauzl packages and types

How to Use

1. Understand Session Protocols

   • Start with AGENTS.md to learn agent session startup, memory conventions, and context continuity processes.

2. Configure and Follow Persona

   • Review SOUL.md to ensure your agent behavior and responses match the designated persona and standards.

3. Maintain Memory Effectively

   • Use MEMORY.md for long-term memory, capturing only events and information directed by the user, and following strict data provenance and verification rules.

4. Customize Local Tools

   • Use TOOLS.md as a template to note local configurations such as device names, SSH hosts, and TTS options.

5. Reference Package Documentation

   • For package information or type definitions, refer to the respective README.md and package.json files under ppt-design/node_modules/.

Notes

• All operational and persona rules are detailed in the Markdown documents at the root level.
• This workspace does not contain runtime code; it provides configuration, conventions, and documentation to structure agent-based environments.
• Package and type documentation under ppt-design/node_modules/ are included for reference and integration with Node.js-based tooling or extensions.
• Files and conventions include both English and Chinese documentation, ensuring clarity for multilingual setups.