forge-convoy

Parallel task coordination across multiple workers is handled by grouping related work into convoys, using a ledger-based pattern to track progress and manage dependencies without direct agent-to-agent communication. This reduces contention and retry complexity in distributed environments.

A convoy pattern for distributed task processing is a coordination model that groups related work bundles with fields like id, tasks, status, and dependencies. It enables parallel execution across workers while preserving audit trails and supporting failure-safe restarts.

Task dependencies and lifecycle states are managed through a convoy data model tracking states from CREATED to IN_PROGRESS, BLOCKED, and COMPLETE. This ledger-based coordination ensures progress tracking and resilient execution across distributed worker restarts.

Yes, you can track progress across distributed workers without direct agent-to-agent communication by using a ledger or beads-style coordination workflow. The convoy model centralizes status updates and dependency management to avoid worker contention.

Failure-safe restarts for parallel task execution are enabled through ledger-based coordination patterns within the convoy model. This approach preserves audit trails and task states, allowing distributed workers to resume coordinated work without losing progress.

You can coordinate multiple Figma components by creating a convoy from a Figma analysis, using fields like figmaFileKey and frameId. This dispatches related component tasks to workers for parallel processing while tracking overall convoy progress.