Does this skill physically block Claude from skipping steps?

Currently, the skill operates as a high-priority guidance specification and progress tracker. While it provides mandatory checkpoints, it relies on Claude's self-compliance and session-end verification rather than hard system-level blocks.

Which workflow steps are considered mandatory?

The default configuration mandates Step 0 (Workflow Preparation), Step 10 (Pre-commit Review), Step 16 (PR Review), and Step 17 (Review Feedback Implementation).

How is the workflow state stored?

State can be managed via the TodoWrite tool or a YAML file located at .claude/runtime/workflow_state.yaml, allowing progress to persist across different parts of a session.

What is the 'Irony' mentioned in the documentation?

It refers to the fact that the skill was created because an agent skipped steps while building it; it acknowledges that AI context decay is the primary problem the skill aims to solve through constant reminders.

Workflow Enforcement

Name: Workflow Enforcement
Author: rysweet

byrysweet

•

Productivity & Workflow

Enforces strict adherence to development workflows by tracking progress and mandating critical checkpoints like code reviews and PR preparation.

About

The Workflow Enforcement skill prevents AI agents from skipping essential steps in the software development lifecycle, such as initial planning, pre-commit reviews, and feedback implementation. By providing visual progress indicators and persistent state tracking via YAML or TodoWrite, it ensures Claude remains accountable throughout a 22-step workflow. This skill mitigates common issues like completion bias and context decay, ensuring that 'PR created' is never confused with 'task completed' without the necessary quality gates.

Key Features

Mandatory gatekeeping for critical steps (Prep, Review, PR Feedback)
Pre-PR validation checkpoints to prevent skipping security audits
17 GitHub stars
Session-end verification via Power Steering integration
State persistence using YAML templates or TodoWrite integration
Visual progress tracking with percentage-based indicators

Use Cases

Ensuring AI agents perform comprehensive code reviews before code is committed
Standardizing development protocols across a team using agentic coding tools
Maintaining process compliance during complex, multi-step architectural migrations

About

Key Features

Mandatory gatekeeping for critical steps (Prep, Review, PR Feedback)
Pre-PR validation checkpoints to prevent skipping security audits
17 GitHub stars
Session-end verification via Power Steering integration
State persistence using YAML templates or TodoWrite integration
Visual progress tracking with percentage-based indicators

Use Cases

Ensuring AI agents perform comprehensive code reviews before code is committed
Standardizing development protocols across a team using agentic coding tools
Maintaining process compliance during complex, multi-step architectural migrations