team-topologies-ai-skill

Prevent coordination chaos by organizing multi-agent work with clear ownership boundaries, defined interaction modes, and bounded cognitive load for each team. This structured governance ensures agents operate within their designated limits without overlapping responsibilities.

The defined interaction modes for multi-agent architecture transitions are stream-aligned, platform, enabling, and complicated-subsystem. Assigning these specific team types establishes structured cross-team collaboration and scalable sub-agent design patterns.

Assign bounded cognitive load for repo planning by restricting each agent's scope of responsibilities to a single domain or subsystem. This prevents overload and maintains clear ownership boundaries throughout the architecture transition.

Yes, this approach supports self-service platform capabilities by designating platform teams to provide underlying infrastructure without direct coordination overhead. This enables stream-aligned agents to autonomously deliver features while maintaining governance boundaries.

You should not use team topologies for multi-agent coordination when your project involves a single agent operating independently or lacks complex cross-team dependencies. The overhead of defining interaction modes is unnecessary for isolated, simple sub-agent work.

The best way to organize a multi-agent project is to assign bounded ownership, interaction modes, and governance boundaries before starting work. This establishes scalable patterns for sub-agent design and prevents misaligned ownership across teams.