Hackathon: Experimenting with new skills despite Docker‐based installation

Using the OmegaClaw Memory Folder for Skills and Code

Even in Docker containers, OmegaClaw has write access to its own memory folder.

For the current setup, that folder appears to the agent as:

./repos/OmegaClaw-Core/memory/

This means the folder can be used not only for stored memories, but also as a lightweight extension space for agent-side resources.

What can be placed there?

You can add:

• Python scripts
• MeTTa code
• Markdown skill descriptions
• notes or instructions the agent should be able to read later

How the agent can use it

The agent can be instructed to:

• read skill descriptions from the memory folder
• invoke tools based on those descriptions
• inspect or execute Python scripts placed there
• load or refer to MeTTa code stored there
• maintain lightweight documentation for its own future operation

Why this matters

This provides a simple way to extend OmegaClaw without changing the provided Docker images.

The memory folder can serve as a minimal agent-readable and agent-writable workspace where operational knowledge, helper scripts, skill descriptions, and MeTTa resources can accumulate over time.

In short:

./repos/OmegaClaw-Core/memory/

is not only storage. If needed, it can also function as a small working directory for skills, scripts, and MeTTa resources.