Where are workflow definitions stored?

Workflows are stored as YAML files within the .claude/workflows/ directory of your project repository.

How do I iterate over multiple tasks in a sprint?

You can use a 'for-each: step' phase in your workflow, which will loop through every step defined in your SPRINT.yaml and apply the specified sub-workflow to each.

Can I run background tasks during a workflow?

Yes, by using per-iteration hooks in Ralph Mode with the 'parallel: true' setting, you can run deterministic tasks like learning extraction or validation in the background.

What template variables are available for workflows?

Common variables include {{step.prompt}}, {{step.id}}, and {{step.index}} for iterations, as well as global variables like {{sprint.id}}, {{sprint.name}}, and {{phase.id}}.

What is 'Ralph Mode' in a workflow?

Ralph Mode enables autonomous, goal-driven execution where Claude dynamically creates and evaluates its own steps based on a goal prompt rather than following a rigid sequence of predefined phases.

Creating Sprint Workflows

Name: Creating Sprint Workflows
Author: Mission42-ai

byMission42-ai

生産性とワークフロー

Defines and manages modular execution paths for automated AI development sprints and task orchestration.

概要

This skill provides a comprehensive framework for authoring sprint workflow definitions within the .claude/workflows/ directory. It enables developers to structure complex tasks into manageable phases, implement 'for-each' logic for step iteration, and leverage 'Ralph Mode' for autonomous, goal-driven execution. By providing standardized YAML schemas and template variable support, it allows for the creation of consistent, repeatable, and highly sophisticated AI development patterns ranging from simple sequential bug fixes to complex feature implementations.

主な機能

Autonomous 'Ralph Mode' for dynamic, goal-oriented task execution
Phase-based sequencing with support for 'for-each' step iteration
Standardized YAML workflow schema for modular development orchestration
Deterministic per-iteration hooks for background tasks and learning extraction
Extensive template variable system for runtime context injection
0 GitHub stars

ユースケース

Automating multi-step QA and documentation workflows across sprint steps
Configuring autonomous agents to solve high-level goals without predefined steps
Standardizing project-specific development pipelines for features and bug fixes

概要

主な機能

Autonomous 'Ralph Mode' for dynamic, goal-oriented task execution
Phase-based sequencing with support for 'for-each' step iteration
Standardized YAML workflow schema for modular development orchestration
Deterministic per-iteration hooks for background tasks and learning extraction
Extensive template variable system for runtime context injection
0 GitHub stars

ユースケース

Automating multi-step QA and documentation workflows across sprint steps
Configuring autonomous agents to solve high-level goals without predefined steps
Standardizing project-specific development pipelines for features and bug fixes