How does this skill prevent AI hallucinations?

It enforces a strict rule set where the AI must only analyze content it has explicitly read and must tag every statement with a certainty level while citing specific file paths and line numbers.

Does this skill help with generating new documentation?

Absolutely. It uses Derivation Tags like [Source], [Derived], and [Generated] to clearly distinguish between original requirements and AI-generated content.

Can I configure the language for these tags?

Yes, the skill supports both English and Traditional Chinese. You can specify your preference in the project's CONTRIBUTING.md file.

What are the primary certainty tags used?

The skill uses five main tags: [Confirmed] for direct evidence, [Inferred] for logical deduction, [Assumption] for patterns needing verification, [Unknown] for missing data, and [Need Confirmation] for user clarification.

AI Collaboration Standards

Name: AI Collaboration Standards
Author: AsiaOstrich

byAsiaOstrich

•

Collaboration Tools

Ensures accurate, evidence-based AI responses by enforcing strict certainty tagging and source verification during code analysis.

The AI Collaboration Standards skill establishes a rigorous framework for AI-human interaction, specifically designed to eliminate hallucinations and ambiguity in software development. By implementing a standardized system of certainty tags (such as [Confirmed] and [Inferred]) and requiring mandatory source citations, it ensures that every AI suggestion or analysis is grounded in verifiable project data. This skill is essential for complex refactoring, reverse engineering, and architectural decision-making where precision and trust in AI-generated insights are critical.

Key Features

01Multi-language support for English and Traditional Chinese tag configurations

02Unified Certainty Tagging System ([Confirmed], [Inferred], [Assumption])

0340 GitHub stars

04Mandatory Source Citation including file paths and line numbers

05Evidence-based analysis rules to prevent AI hallucinations

06Content derivation tracking for generated specifications

Use Cases

01Generating technical specifications directly derived from project requirements

02Analyzing legacy codebases where architectural certainty is required

03Conducting AI-assisted code reviews with verifiable evidence-based feedback

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add asiaostrich/universal-dev-standards ai-collaboration-standards

For use in Claude.ai and ChatGPT

Download Skill