How does it handle context window limits?

The skill monitors token usage, triggering predictive cleanups at 80% and automated commits or checkpoints at 90% to ensure session continuity and prevent data loss.

Does it support automated testing?

Yes, it includes built-in phases for E2E validation via Playwright and unit test verification via Pytest, including automated coverage reporting and debugging of failed tests.

What is the 'Ralph Wiggum' philosophy in this skill?

It is the core principle that the agent should never idle; if no explicit tasks are assigned, the engine proactively discovers improvements like fixing linting errors, increasing test coverage, or updating documentation.

How do I start an autonomous session?

You can initiate the loop using the included auto_cli.py script, which allows for infinite loops, max iteration limits, or specific completion promises.

Which Claude commands does it automate?

It automatically triggers /debug, /check, /commit, /issue, /pr, /tdd, /research, /audit, and /parallel based on its internal 5-tier priority logic.

Autonomous Workflow Engine

Name: Autonomous Workflow Engine
Author: garimto81

bygarimto81

•

Productivity & Workflow

Orchestrates a self-correcting, multi-tiered autonomous development loop that proactively identifies and executes tasks.

The Autonomous Workflow Skill (auto-workflow) transforms Claude Code from a reactive tool into a proactive agent capable of independent discovery and task execution. Built on the 'Ralph Wiggum' philosophy—where the absence of a task triggers autonomous discovery rather than termination—it manages a 5-tier priority system ranging from critical session management to long-term architectural improvements. It integrates nine specialized commands (debug, check, commit, etc.), automates context management to prevent token overflow, and enforces rigorous E2E and TDD validation, making it indispensable for large-scale automation and continuous codebase maintenance.

Key Features

015-Tier Priority System (Emergency to Autonomous Discovery)

02Ralph Wiggum 'Self-Discovery' Mode for continuous improvement

039-Command Auto-Triggering for seamless dev loops

041 GitHub stars

05Predictive Context Management with auto-cleanup at 90%

06E2E and TDD Validation with integrated Playwright and Pytest

Use Cases

01Hands-free debugging and error resolution using a multi-agent parallel approach.

02Proactive feature development where the agent analyzes PRDs and handles the full TDD lifecycle.

03Automated maintenance of large-scale codebases through continuous linting and security patching.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add garimto81/claude auto-workflow

For use in Claude.ai and ChatGPT

Download Skill