team-workload

Team Workload

Analyze agent capacity, detect overload conditions, and rebalance conversation assignments to keep the team running smoothly.

Workflow Modes

View Workload (default)

Display the current workload distribution across the entire team.

1. Fetch team members using list_team_members to get all agents, their roles, and online status
2. Fetch open conversations using list_conversations with status: "open" to see all active work
3. Count conversations per agent by grouping on the assignee field
4. Count unassigned conversations separately
5. Fetch agent performance data using get_agent_performance for each agent to add response time and resolution rate context
6. Calculate the distribution ratio: each agent's share of total open conversations as a percentage
7. Flag agents who exceed the overload threshold (12+ conversations)
8. Present the workload table

Rebalance Queue

Redistribute conversations from overloaded agents to those with available capacity.

1. Run the View Workload workflow to get current state
2. Identify overloaded agents (12+ conversations) and underloaded agents (fewer than 6 conversations)
3. For each overloaded agent, select conversations to reassign — prefer newer, lower-priority, and unresponded conversations
4. Build a rebalance plan: which conversations move from which agent to which agent
5. Present the rebalance plan to the user for confirmation
6. On confirmation, execute reassignments using assign_conversation for each conversation
7. Also assign any unassigned conversations to agents with the most available capacity
8. Present a before-and-after summary showing the new distribution

Check Specific Agent

Show detailed workload and performance data for a single agent.

1. Fetch team members using list_team_members to find the named agent
2. Fetch open conversations using list_conversations with status: "open" and filter to those assigned to the agent
3. Fetch the agent's performance data using get_agent_performance
4. Optionally fetch overall metrics using get_metrics for comparison
5. Present the agent detail card with their conversations, performance stats, and capacity status