Can I use this for English-speaking projects?

Yes. Although this version includes Chinese localization support, the standards are universal and the tags default to English unless configured otherwise in your project's contributing guidelines.

How does this skill prevent AI hallucinations?

It mandates that the AI only provide statements based on explicit evidence found in the code or documentation, using certainty tags to label the reliability of every statement.

Why are source references mandatory in this skill?

Source references provide a direct audit trail to specific file paths and line numbers, allowing human developers to quickly verify the AI's claims against the actual codebase.

What are certainty tags?

Certainty tags like [Confirmed], [Inferred], and [Assumption] are labels that the AI uses to communicate its level of confidence and the nature of its reasoning for every response.

AI Collaboration Standards

Name: AI Collaboration Standards
Author: AsiaOstrich

byAsiaOstrich

•

协作工具

Ensures accurate, evidence-based AI responses by implementing standardized certainty tagging and source referencing protocols.

This skill establishes a rigorous framework for AI-human collaboration designed to eliminate AI hallucinations during the development process. By enforcing the use of certainty tags—such as [Confirmed], [Inferred], and [Assumption]—and requiring mandatory source citations with file paths and line numbers, it ensures that every AI response is verifiable and transparent. It is an essential tool for teams who require high-integrity code analysis, architectural decision-making, and reliable documentation within their software projects.

主要功能

01Anti-hallucination protocols based strictly on explicit evidence

02Mandatory source referencing for all code and documentation claims

0340 GitHub stars

04Bilingual support for tag localization in English and Chinese

05Structured recommendation engine for presenting technical options

06Standardized certainty tagging (Confirmed, Inferred, Assumption, Unknown)

使用场景

01Debugging complex logic by tracing AI claims to specific line numbers

02Drafting architectural documentation based on logical inference from existing code

03Conducting verifiable code reviews with clear evidence markers

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