refine-skills

Extract actionable feedback from skills used in the current session.

Updated Mar 31, 2026
One-click install
npx skills add https://github.com/Tanguyvans/claude-config --skill refine-skills
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: refine-skills
Source: https://github.com/Tanguyvans/claude-config/tree/main/skills/refine-skills
Command: npx skills add https://github.com/Tanguyvans/claude-config --skill refine-skills

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Analyze the current conversation to extract actionable feedback for the skills used, helping you improve future interactions and session outcomes.

Core Features & Use Cases

  • Identify calls to /skill-name and analyze outcomes for each used skill.
  • Capture issues, user corrections, frustrations, and missing steps.
  • Provide categorized feedback (improvement, positive, bug, missing) and support validation workflows.

Quick Start

Run refine-skills to generate actionable feedback for all skills used in the current session.

Frequently Asked Questions about refine-skills

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract actionable feedback from skills used in a conversation?

Capture conversation feedback by identifying skill calls and analyzing outcomes to surface issues, user corrections, and missing steps. The feedback is categorized into improvement, positive, bug, or missing types and validated through user confirmation workflows.

What is the best way to automate skill improvement analysis after a session?

Automate skill improvement analysis by running a feedback extraction process across the conversation to identify frustrations and missing steps. It validates the extracted feedback through user confirmation before storing it in a global repository or local project scope.

Can I categorize and validate conversation feedback for specific skill invocations?

Yes, you can categorize conversation feedback into improvement, positive, bug, and missing categories for specific skill invocations. The validation workflow requires user confirmation to ensure accuracy before the feedback is stored in a repository.

Does skill feedback validation support both global repo and local project scope storage?

Yes, skill feedback validation supports both global repo and local project scope storage. After user confirmation validates the categorized feedback, the results can be stored in either scope to help improve future interactions and session outcomes.

How do I identify missing steps and user corrections from previous skill calls?

Identify missing steps and user corrections by analyzing the outcomes of previous skill calls within the current conversation. The extraction process surfaces these issues automatically and categorizes them as actionable feedback for validation.