openclaw-workspace

Summary

Templates and guides for building and managing complex AI-driven workspaces in OpenClaw. Includes agent profiles, memory architecture, tool configuration, onboarding flows, project examples, and skill guides.

Included Assets

Top-level files:

• AGENTS.md — Agent operating procedures & workspace rules
• MEMORY.md — Long-term memory and user context
• SOUL.md — Agent personality, work/casual modes
• TOOLS.md — Local environment notes

Project folders:

• projects/brutanka-novel/ — Debate engine for fiction, agents, scenarios (README.md, agent SOULs)
• projects/job-hunting/ — Job search workflows for Eliza and Sergiy (README.md files)
• projects/onboarding/ — New user onboarding flow and message templates

Skills:

• skills/humanizer/ — Text naturalization skill (README.md)
• skills/music-api-ai/ — Music AI API integration (README.md, scripts, workflow)
• skills/self-improving-agent/assets/ — Agent, memory, soul, and tools asset templates

Memory:

• Daily and batch memory logs (memory/2026-03-13.md, memory/2026-03-13-jobs-batch.md, memory/2026-03-15.md)

How to Use

1. Agents: See AGENTS.md, SOUL.md for setup, session workflow, and agent identity rules. Adapt to your workspace needs.
2. Memory: Store daily logs in memory/YYYY-MM-DD.md. Curate key facts in MEMORY.md. Use these for continuity and user context.
3. Tools: Document local hardware, SSH, device names, and environment-specific notes in TOOLS.md.
4. Skills: Review skill folders for install and usage guides. Integrate humanizer, music-api-ai, or agent improvement assets as needed.
5. Projects: Follow each README.md under projects/ for domain-specific flows (job hunting, onboarding, debate simulation).
6. Self-improving agent: Use asset templates in skills/self-improving-agent/assets/ to design proactive, resourceful agents.

Notes

• All templates provided are fully editable; adapt them to match your workspace and user profile.
• Session and memory logging guidance is detailed in AGENTS.md.
• Project folders demonstrate real-world use cases; clone or extend for your own applications.
• Skill modules are written in Python and Markdown for easy integration.
• Memory logs are structured for continuity; curate or purge as needed.
• This workspace is focused on customizable agent-driven workflows, not on executable workflows or automated orchestration.