How does this skill prevent Claude from skipping steps?

It uses auto-activation triggers and mandatory checkpoints that require Claude to verify the completion state of previous steps before proceeding to high-level actions like opening a Pull Request.

What happens if a mandatory step is missed?

The skill provides explicit blocking instructions and visual warnings at defined checkpoints (like Step 15 and 21), prompting the agent to return and complete the missing requirement.

Where is the workflow state stored?

State is tracked either through the built-in TodoWrite tool or a specialized workflow_state.yaml file located in the .claude/runtime directory.

Does this actually block Claude from making a PR?

In its current state, it acts as high-authority guidance and self-compliance logic. Hard-blocking requires integration with external pre-commit hooks or CI checks.

Can I customize the workflow steps?

Yes, the skill is designed to align with a project's DEFAULT_WORKFLOW.md file, which can be modified to suit your specific team requirements and mandatory gates.

Workflow Enforcement for Claude Code

Name: Workflow Enforcement for Claude Code
Author: rysweet

byrysweet

•

生产力与工作流

Enforces standardized agentic coding workflows through mandatory step validation and visual progress tracking.

关于

This skill optimizes the reliability of AI-driven development by guiding Claude through a rigorous 22-step workflow designed for agentic coding. It mitigates 'completion bias'—the tendency for AI agents to skip mandatory code reviews or PR steps—by implementing persistent state tracking, visual progress indicators, and hard checkpoints. By defining mandatory gates at preparation, review, and feedback stages, it ensures that every task from initial workspace setup to the final merge follows a documented, repeatable process that maintains high code quality standards.

主要功能

Integration with session-end verification to ensure full process compliance
Auto-activation triggers for common development workflow start sequences
Mandatory gate enforcement for critical steps like security reviews and PR feedback
17 GitHub stars
Visual progress bars and detailed status dashboards for real-time monitoring
Persistent workflow state tracking via YAML or TodoWrite integration

使用场景

Ensuring AI agents don't skip mandatory security and code review steps before PR creation
Tracking progress across complex, multi-session engineering tasks to prevent context decay
Standardizing PR preparation and feedback implementation across distributed developer teams

关于

主要功能

Integration with session-end verification to ensure full process compliance
Auto-activation triggers for common development workflow start sequences
Mandatory gate enforcement for critical steps like security reviews and PR feedback
17 GitHub stars
Visual progress bars and detailed status dashboards for real-time monitoring
Persistent workflow state tracking via YAML or TodoWrite integration

使用场景

Ensuring AI agents don't skip mandatory security and code review steps before PR creation
Tracking progress across complex, multi-session engineering tasks to prevent context decay
Standardizing PR preparation and feedback implementation across distributed developer teams