How does the complexity triage system work?

It categorizes tasks into Quick, Standard, or Deep paths. Simple edits get a fast check, while core system or security changes require a deep reflection protocol and high confidence scores.

How does it handle performance claims?

The skill includes a mandatory fact-checking step where any claims regarding speed, memory, or complexity must be verified against documentation or through algorithmic analysis.

What is the primary purpose of the Reflexion skill?

It acts as a quality gatekeeper that forces the AI to critically evaluate its own work, looking for logical flaws, missing requirements, and architectural anti-patterns.

Can this skill help reduce technical debt?

Yes, by enforcing a 'Library-First' approach and checking for SOLID principles, it prevents the creation of redundant custom utilities and poorly structured code.

Does it verify external dependencies?

Yes, the framework includes a specific protocol to search for existing libraries (npm, PyPI, etc.) and verify that proposed changes don't break existing ecosystem dependencies.

Reflexion Self-Refinement Gatekeeper

Name: Reflexion Self-Refinement Gatekeeper
Author: NeoLabHQ

byNeoLabHQ

•

542

•

Developer Tools

Implements a ruthless self-correction framework to catch flaws, verify dependencies, and ensure production-grade code quality.

The Reflexion skill transforms Claude into a critical quality gatekeeper that subjects its own outputs to a rigorous multi-stage verification process. By applying complexity triage—ranging from Quick Path to Deep Reflection—the skill forces the model to identify architectural anti-patterns, verify external dependencies, and fact-check technical claims before finalizing work. It is specifically designed to eliminate 'Not Invented Here' syndrome by prioritizing established libraries and ensuring all code aligns with Clean Architecture, SOLID principles, and domain-driven design.

Key Features

01Fact-Checking Protocol: Mandates verification for performance claims, security assertions, and technical facts.

02542 GitHub stars

03Complexity Triage: Automatically scales reflection depth based on the task's architectural impact.

04Library-First Enforcement: Actively searches for and suggests established open-source solutions to prevent reinventing the wheel.

05Ruthless Quality Gating: Subjects outputs to a perfectionist critique to catch deficiencies before they ship.

06Dependency & Impact Verification: Validates proposed changes against existing system configurations and documentation.

Use Cases

01Auditing complex feature implementations for architectural alignment and code smells.

02Ensuring new code correctly utilizes existing ecosystem dependencies without duplication.

03Reviewing critical core system changes where a single error could cause production downtime.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add neolabhq/context-engineering-kit reflect

For use in Claude.ai and ChatGPT

Download Skill