research-orchestration

Multi-wave research orchestration coordinates tasks across separate cohorts by using a centralized orchestrator to dispatch sequential waves and route work to companion skills. This ensures consistent outputs and provenance across all cohorts.

The wave model applies a predefined four-wave sequence (Wave 1 through Wave 4) to structure research tasks. It plans and dispatches waves sequentially to ensure repeatable research outputs and maintain independence across cohorts.

Research orchestration provides an orchestration contract that automatically routes specific tasks to companion skills like academic-writing, source-evaluation, and web-scraping-foundations. This routing happens dynamically as the orchestrator dispatches each wave.

You need multi-cohort orchestration when splitting large research work across independent agents to maintain provenance and prevent cross-contamination. It isolates cohorts to ensure structured, repeatable research with consistent standardized outputs.

Trigger research orchestration at project start to initialize the wave plan and companion-skill routing. The orchestrator schedules the predefined waves and establishes the standardized project tree for storing outputs.

The orchestrator stores outputs in a standardized project tree designed for cross-cohort synthesis and reporting. This standardized placement ensures that all wave and cohort data remains organized and accessible for final synthesis.