openclaw-workspace

Summary

A structured OpenClaw workspace designed for personal agent-based automation, memory management, and workflow organization. Includes agent specifications, persistent and daily memory, tool/environment notes, and specialized sub-projects for dinner recommendations and interview prep tracking.

Included Assets

• Workspace Modules

  • AGENTS.md — Workspace rules, agent session routines, memory guidance
  • MEMORY.md — Long-term, curated personal memory (in Chinese)
  • SOUL.md — Agent identity, operating principles, safety rails
  • TOOLS.md — Local environment/device notes & OpenClaw tool gotchas

• Memory System

  • memory/YYYY-MM-DD.md — Daily logs, events, reminders (dated)

• Agents

  • agents/alpha/AGENT.md — Investment research agent specification
  • agents/investment-alphaforge/AGENT.md — Legacy alias for alpha agent

• Skills

  • skills/humanizer/README.md — Skill for naturalizing AI text
  • skills/proactive-agent/assets/ — Proactive agent module with assets for session rules, memory, soul, and tool configuration

• Apps/Subprojects

  • byte-doubao-prep-tracker/ — Byte Doubao interview preparation tracker (with README and training guides)
  • dinner-what-to-eat/ — Family dinner recommendation webapp (with README and package.json)
  • nodejs-hello-login/ — Simple password-protected hello world web server (Node/Express, with package.json)

How to Use

1. Workspace Bootstrapping

   • Read AGENTS.md for session routines and workspace navigation.
   • Review SOUL.md to understand agent identity and safety measures.
   • Reference MEMORY.md and daily memory/YYYY-MM-DD.md files for personal context, ongoing projects, and user preferences.
   • Use TOOLS.md to document and reference local hardware, preference, or environment-specific information.

2. Agents

   • Explore agents/alpha/AGENT.md for investment research agent workflow.
   • Maintain or update agent memory and behavior descriptions as needed.

3. Apps and Skills

   • See dinner-what-to-eat/README.md for set up and usage of the dinner recommendation app.
   • See byte-doubao-prep-tracker/README.md for structured interview training workflow.
   • Install and use skills (e.g., Humanizer, Proactive Agent) per their documentation.

4. Personal Memory and Routines

   • Add new daily logs in memory/YYYY-MM-DD.md, review recent logs for continuity.
   • Curate key information into MEMORY.md (avoiding sensitive or unapproved data for security).
   • Follow routines defined in AGENTS.md for each session.

Notes

• All core agent routines, safety practices, and operating boundaries are explicitly described in SOUL.md and AGENTS.md.
• Daily to long-term memory distinction is crucial: logs go into dated memory files; only key, non-sensitive info is summarized in MEMORY.md.
• Tool and device information should be added in TOOLS.md to avoid environment-specific leaks in modules/skills.
• Subprojects (dinner-what-to-eat, byte-doubao-prep-tracker, etc.) each contain their own instructions and should be set up individually if needed.
• Security, privacy, and explicit user confirmation are mandatory for all sensitive operations.
• This workspace is designed for hands-on, iterative use with a focus on real personal productivity, not for general public deployment.