What does the confidence score represent?

A score from 0.0 to 1.0 indicates how well the prompt is structured. 0.8-1.0 means all ambiguities are resolved, while lower scores indicate missing critical information.

What is the TCRO framework used in this skill?

TCRO stands for Task (the job to be done), Context (the why and broader system), Requirements (constraints), and Output (the expected deliverable format).

When is the Prompt Refinement skill invoked?

It is automatically triggered at the start of /ai-eng/research, /ai-eng/specify, /ai-eng/plan, and /ai-eng/work commands.

Does this skill require a CLAUDE.md file?

The skill attempts to load CLAUDE.md to understand your project's tech stack and conventions, but it can proceed with generic defaults if the file is missing.

AI Prompt Refiner & TCRO Structurer

Name: AI Prompt Refiner & TCRO Structurer
Author: v1truv1us

byv1truv1us

•

Productivity & Workflow

Transforms vague user input into highly structured, TCRO-formatted prompts to maximize AI output quality and consistency.

The Prompt Refinement skill is a core utility for the ai-eng-system designed to eliminate the ambiguity that leads to poor AI performance. It implements the TCRO (Task, Context, Requirements, Output) framework to systematically structure user requests. By utilizing phase-specific templates—Research, Specify, Plan, and Work—the skill asks targeted questions to uncover hidden constraints and project-specific requirements. It integrates directly with project context via CLAUDE.md and applies advanced incentive-prompting techniques like expert persona assignment and stakes language to ensure Claude delivers high-fidelity, production-ready results while providing a confidence assessment of the final prompt.

Key Features

01Automated integration with project-specific standards in CLAUDE.md

02Incentive-prompting enhancement including expert personas and step-by-step reasoning

032 GitHub stars

04TCRO Framework implementation (Task, Context, Requirements, Output)

05Confidence assessment scoring (0.0-1.0) to evaluate prompt readiness

06Phase-specific clarification templates for different development stages

Use Cases

01Converting brief feature requests like 'need auth' into full technical specifications

02Structuring complex research tasks to ensure comprehensive technology evaluations

03Refining implementation plans to reduce iteration cycles and catch edge cases early

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add v1truv1us/ai-eng-system prompt-refinement

For use in Claude.ai and ChatGPT

Download Skill