improve-team

To optimize multi-agent team performance, you apply targeted, structure-preserving edits to team rosters and collaboration patterns based on execution evidence. This adjusts operational thresholds and resolves bottlenecks without requiring a full system rewrite.

Continuous agent teams experience high token spend and oscillating results when operational constraints are misaligned. You can tune the budget, tighten verifier criteria, or promote the team's maturity level to improve autonomy and reliability based on run evidence.

The best way to adjust collaboration patterns safely is through structural refinement. This approach applies targeted edits to serial, parallel, or continuous collaboration workflows based on execution metrics, preserving the overall system structure.

Yes, you can add dedicated agents to existing team rosters to resolve territory conflicts. Evidence-driven optimization identifies overlapping responsibilities and supports targeted structural edits to refine team composition.

Yes, you need run reports and feedback logs to tune operational thresholds. Evidence-driven optimization maps performance metrics to specific structural changes, identifying convergence issues and bottlenecks for targeted adjustments.

You should avoid structural refinement for agent management when a full system rewrite is required. This approach relies on targeted, structure-preserving edits, making it unsuitable for fundamentally redesigning team architectures or completely replacing collaboration frameworks.