$ loading_
Orchestrate multi-agent software engineering workflows with parallel workers and code integration.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "Cohort" yet — see the docs or source repo.
Create a bespoke software engineering organization for “adding user role and permission management to an existing web app,” break down tasks, run parallel workers in isolated git worktrees, and integrate reviewed code through a replan loop.
A multi-agent plan, parallel implementation workflow, and reviewed integrated code changes.
Set up a multi-agent engineering workflow for “refactoring the authentication module and reducing coupling,” let workers handle subtasks in parallel in separate worktrees, and merge changes after review-driven replanning.
A phased refactoring workflow and integrated code updates refined through review and replanning.
Development teams can use it to generate a bespoke multi-agent organization for large software engineering objectives. It splits work and coordinates parallel execution to reduce manual orchestration effort.
When multiple subtasks need to move forward at the same time, it can run parallel workers in isolated git worktrees. This helps reduce direct conflicts and supports later review and integration.
It fits engineering workflows where plans need to evolve during implementation. Through a replan loop, it integrates reviewed code so execution can adapt based on intermediate results.
Cohort is an autonomous multi-agent orchestration layer that generates a bespoke software engineering organization for a given objective. It runs parallel OpenCode workers and handles development and integration in isolated git worktrees.
Based on the description, it integrates code after review through a replan loop. In other words, integration is tied to both review and replanning.
The provided material does not include installation steps, dependencies, or key requirements. See the source repository for details.
Guide coding agents through structured, evidence-based software engineering with persistent state.
Coordinate multi-agent teamwork with ownership, Kanban flow, merge gates, and handoffs.
Coordinate coding agents locally with shared memory, tasks, and handoffs.
Maintain append-only shared memory and review-based acceptance for collaborative work.
Orchestrate multi-agent workflows with parallel tasks, pipelines, scheduling, and peer review.
Plan and orchestrate parallel coding agents with progress tracking and reports.