skills.rest
sizukutamago
sizukutamagoCOMMUNITY
Explore all skills by this author →
GitHub

prompt-improver

Collect task completion feedback and store it as structured YAML.

Updated Feb 4, 2026
Search Tags:#workflow#feedback#CLAUDE.md#analyze#improve#Stop hook#prompt-improver
One-click install
npx skills add https://github.com/sizukutamago/dev-tools-plugin --skill prompt-improver-sizukutamago
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-improver
Source: https://github.com/sizukutamago/dev-tools-plugin/tree/main/skills/prompt-improver
Command: npx skills add https://github.com/sizukutamago/dev-tools-plugin --skill prompt-improver-sizukutamago

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python3, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

Collects feedback on task completion and stores it as structured YAML to enable continuous prompts improvement, CLAUDE.md updates, and skill refinements.

Core Features & Use Cases

  • Automatic Stop-hook feedback collection after task completion.
  • Analysis of collected feedback to identify patterns and hotspots.
  • Generation of concrete improvement proposals for prompts, CLAUDE.md, and skills.
  • Interactive application flow to apply improvements and update triage data.
  • Seamless integration with existing scripts, references, and assets for guided enhancements.

Quick Start

Install the prompt-improver skill with the provided install.sh, then run tasks and use the /improve command to generate improvements.

Frequently Asked Questions about prompt-improver

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

FAQPage Schema
How do I collect feedback automatically to improve prompts and CLAUDE.md configurations?▼

To improve prompts and CLAUDE.md configurations, you can use a Stop hook to collect feedback automatically after task completion. The skill stores this feedback as structured YAML, enabling continuous analysis and actionable updates without manual input.

What is the best way to analyze task completion feedback for prompt improvement?▼

The best way to analyze task completion feedback is to store it as structured YAML and run optional Python extraction scripts. This identifies patterns and hotspots, generating concrete improvement proposals for your prompts and skills.

Do I need Python to use a Stop hook for collecting prompt feedback?▼

You do not need Python to use a Stop hook for collecting prompt feedback. The setup requires only standard Unix tools, while Python3 is an optional dependency used for advanced extraction and analysis workflows.

How do I generate concrete improvement proposals from collected prompt feedback?▼

You generate concrete improvement proposals from collected prompt feedback by running an interactive application flow. This analyzes the stored YAML data and provides actionable recommendations to update prompts, CLAUDE.md, and skills.

Can I update CLAUDE.md and skills interactively based on structured feedback?▼

Yes, you can update CLAUDE.md and skills interactively based on structured feedback. The workflow applies generated improvements and updates triage data seamlessly, integrating with existing scripts and references for guided enhancements.

Related Skills

Explore More Content & Communication Skills→
89jobrien89jobrien
4

claude-hooks

Configure Claude Code hooks for workflow automation and audit logging.

Community
Intermediate
paulmalmquistpaulmalmquist

winston-prompt-normalization

Scan Winston prompt docs, classify prompts, and update skill contracts and routing references.

Community
Intermediate
korchasakorchasa
1

flowai-skill-engineer-command

Create flowai commands with YAML frontmatter and standard resource layouts.

Community
Advanced
gtappsgtapps
69

reflect

Analyze recent work and cost logs to generate structured improvement proposals.

Community
Advanced