sensemaking-concentrator

Audit multi-agent system sensemaking conflicts by mapping cross-agent signal flows to identify read and write overlaps. The skill requires a declarative model of agents and signals to detect interpretation anti-patterns and recommend concentration strategies.

Common sensemaking anti-patterns occur when multiple agents interpret the same input differently, generating conflicting signals. The skill identifies these overlaps and recommends concentrating interpretation into a single owner to improve decision quality.

Consolidate sensemaking by applying concentration strategies like designating a single owner, deploying a concentrator agent, or establishing an escalation protocol. The skill outputs concrete implementation steps, interface requirements, and context specifications.

Yes, a declarative model of agents and signals is required to map cross-agent signal flow. The skill uses this model to identify read and write overlaps, pinpointing zones with potential sensemaking conflicts and concentration opportunities.

Debug conflicting agent actions by auditing the distributed architecture to map signal flows. The skill identifies zones where agents propose divergent actions from identical inputs and recommends unified concentration strategies.

Concentrate interpretation into a single agent when cross-agent signal mapping reveals read and write overlaps causing conflicting outputs. The skill details specific zones, owners, and implementation steps for resolving these distributed sensemaking conflicts.