What problem does it solve? Coordinating multiple distributed funboost tasks that depend on each other—running steps in sequence, in parallel, or aggregating parallel results into a callback—requires manual glue code. This Skill provides Celery-style primitives (chain, group, chord) to compose multi-step task pipelines declaratively. ## Core Features & Use Cases - Serial Pipelines with chain: Pass each task's return value as the first argument to the next task, e.g., download -> process -> upload. - Parallel Execution with group: Run multiple task signatures concurrently, such as transcoding one video into 360p, 720p, and 1080p at the same time. - Fan-out/Fan-in with chord: Execute a group in parallel, then collect all results as a list into a callback task like a user notification. - Nested Composition: Combine chain, group, and chord arbitrarily to model complex multi-stage workflows. - Use Case: Build a video processing pipeline where a download task feeds a chord that transcodes in parallel and finally notifies the user with the aggregated results. ## Quick Start Ask the AI to write a funboost workflow that chains a download task into a chord of parallel processing tasks with a notification callback, using WorkflowBoosterParams and calling consume() on every task.
