openclaw-workspace

Summary

This workspace organizes and manages configuration, documentation, and operational history for OpenClaw agents, memory, and tool integrations. It provides strict operational guidelines, tooling notes, and historical session records to support robust, reliable agent functionality in the OpenClaw ecosystem.

Included Assets

• AGENTS.md — Agent behavioral directives, non-negotiable rules for response integrity and user interactions.
• SOUL.md — Defines the agent's personality and general conduct principles.
• MEMORY.md — Guidance and policy for accessing and managing networking devices and related memory.
• TOOLS.md — Documentation of system skills (tool interfaces) and their usage patterns.
• memory/ — Date-stamped logs and session summaries detailing past agent operations, events, and audit trails.

How to Use

1. Review Operational Rules: Start with AGENTS.md for agent response protocols—especially the strict policies on data sourcing and user interaction.
2. Understand Persona: Reference SOUL.md to align responses and actions with the defined operator style and privacy standards.
3. Leverage Memory and Tools:
   • Consult MEMORY.md for instructions on handling networking devices and correct skill invocation.
   • Use TOOLS.md as a reference for calling system-integrated skills (typically via shell exec commands).
4. Access Historical Logs: Explore the memory/ directory for chronological records of commands, automation, session notes, error reports, and operational adjustments.
5. Maintain Accuracy: Always use defined tools and skills for current system information; do not fabricate or infer device status or measurements.

Notes

• All status or measurement data must be retrieved in real-time from the defined tool calls—never from memory or prior observations.
• Home Assistant, UniFi Network, and UniFi Protect each require their own dedicated skill calls for device or camera data.
• The memory log files in memory/ serve as both an audit of automation and a journal for configuration or troubleshooting steps.
• Any new tools, devices, or system rules should be documented in the relevant .md configuration file within the workspace.
• Follow agent guidelines strictly—failure to comply with non-fabrication rules is considered a critical error.