📖About & Usage

About

Agent Engagement is a read-only dashboard for seeing how Carol’s internal AI agents interact with Carol. It brings together agent conversations in one place, helping administrators understand what the agents have been discussing, how work unfolded, and where follow-up may be useful. Think of it as a window into the team’s working conversations, without any buttons for changing them.

The app is the agent-focused counterpart to User Engagement, which covers conversations with human users. Access is restricted to administrators because agent conversations are confidential. The app is owned by Carol, Carol, and sits within Carol’s monitoring tools. It exists to make agent activity easier to review without requiring someone to piece together the story across separate systems.

Usage Patterns

Agent Engagement is used when an administrator needs to inspect an agent’s conversation history with Carol. It does not send messages, edit records, or direct the agents; it simply presents what happened. This makes it useful for routine oversight, investigating unexpected behaviour, checking whether instructions were understood, and reconstructing the context behind an agent’s decision.

For example, suppose Argus, Argus, reports an unusual test result. An administrator can open Agent Engagement, locate Argus’s relevant conversations, and follow the exchange that led to the result. They might then compare the surrounding operational picture in Carol Monitor or consult Agent Records for information about the agent itself. Agent Engagement provides the conversational trail—the friendly “show me how we got here” view—while those other apps cover their own parts of the Carol ecosystem.

🗂️Tabs & Screens

Tab inventory is being built — see CAROL-INI-077 step 7.

👤Owner

📚Recent initiatives

Initiatives that touched this app — a short summary each; open one for the full story.