What happens when all acceptance criteria are met?

Once the validator confirms all checklist items have passed, the engine automatically stages changes with 'git add -A', creates a structured commit message, and updates the task status to completed.

Where does the engine store its progress?

It maintains a 'checklist.md' file in the task directory with markdown checkboxes and updates a 'meta.yaml' file with a summary of the current progress (e.g., 3/5 tasks completed).

How does the Fix Engine handle implementation failures?

The skill manages a retry loop (up to 3 attempts). If the validator identifies failures, the engine passes the accumulated failure context back to the executor to refine the implementation.

Can I see why a task failed after the maximum retries?

Yes, if the retry limit is hit, the engine generates a detailed YAML error report containing the section ID, history of approaches taken, specific code snippets, and tested methods for escalation.

InfiStack Fix Engine

Name: InfiStack Fix Engine
Author: blueif16

byblueif16

0•

Productivity & Workflow

Orchestrates an automated executor-validator loop to implement features, verify acceptance criteria, and handle retries with systematic checklist tracking.

The InfiStack Fix Engine is a specialized Claude Code skill designed for sub-coordinator agents to manage the end-to-end lifecycle of a coding task. It translates requirements from a mini-PRD into a dynamic Markdown checklist and manages a closed-loop execution cycle between implementation and validation agents. The skill handles up to three retry attempts with cumulative error context, ensures all acceptance criteria are met before performing structured Git commits, and generates comprehensive diagnostic reports for escalation if tasks cannot be completed automatically.

Key Features

010 GitHub stars

02Structured Git commit automation upon successful task completion

03Real-time progress tracking via checklist.md and meta.yaml updates

04Detailed error reporting and escalation routing for persistent failures

05Closed-loop executor-validator orchestration with retry logic

06Automated checklist generation from mini-PRD acceptance criteria

Use Cases

01Ensuring high-quality code delivery through systematic acceptance criteria testing

02Automating the bug-fix and verification cycle with cumulative context

03Implementing complex features that require multiple validation passes

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add blueif16/amazing-claude-code-plugins fix-engine

For use in Claude.ai and ChatGPT

Download Skill