Agent Workspace Coordinator

Overview

This agent acts as a central coordinator for defining and enforcing operational protocols across different interaction environments. It dictates the 'role' an AI should adopt—whether it is in a public, shared setting or a private, deep-dive session. Its core function is to maintain context integrity by strictly limiting data access based on the channel type.

Capabilities

• Session Type Detection: Automatically identifies if the current interaction is Public/Shared, Main (Private), or Research Mode.
• Role Assignment: Switches operational persona (e.g., 'Friend' Agent vs. 'Librarian' Agent) based on context.
• Data Gatekeeping: Enforces strict rules regarding private data access (USER_PRIVATE.md, MEMORY.md) in public settings, preventing accidental leakage.
• Memory Retrieval: Manages calls to specific memory modules (e.g., neural-memory.nmem_recap).

Example Use Cases

• Public Discussion: When operating in a group chat, this agent ensures it only uses 'Spirit Only' knowledge, preventing access to sensitive project details.
• Deep Work Session: In a private DM, it grants full access to the user's complete memory and context for comprehensive assistance.
• Project Handover: It coordinates between different agents by acting as an A2A Coordinator, ensuring that summaries passed between 'Friend' agents are safe and filtered.