Prompt Improver

Improve and optimize prompts for Claude using Anthropic's official best practices and the 5-Element Persona Framework. Use when asked to improve, refine, optimize, or review a prompt. Also triggers on requests to make prompts clearer, more effective, better structured, or to get better/more realistic outputs from Claude. Covers Claude 4.x-specific techniques, persona engineering, constraint-based prompting, and structured outputs.

46 stars 56 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add zocomputer/skills --skill prompt-improver
Manual install
git clone --depth 1 https://github.com/zocomputer/skills.git /tmp/skills
mkdir -p ~/.claude/skills
cp -r /tmp/skills/Community/prompt-improver ~/.claude/skills/prompt-improver

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Prompt Improver

Analyze and improve prompts using Anthropic's official best practices and the 5-Element Persona Framework.

## Workflow

### Step 1: Identify the Prompt & Intent
- Get the prompt if not provided
- Clarify: Is this for Claude 4.x? API or chat interface?
- Understand: What outcome does the user want to improve?

### Step 2: Diagnose the Core Issue
Quick triage against common failure modes:

| Symptom | Likely Cause |
|---------|--------------|
| Generic/shallow output | Missing persona depth or constraints |
| Wrong format | No output specification or examples |
| Inconsistent results | Needs few-shot examples with reasoning |
| Unrealistic suggestions | Missing constraints (budget/time/resources) |
| Ignores instructions | Needs XML structure or system/user separation |
| Verbose/over-engineered | Missing explicit direction (Claude 4.x issue) |
| Off-topic responses | Lacks context about purpose/audience |

### Step 3: Apply the 5-Element Persona Framework
For any prompt involving expertise or role-play, ensure ALL five elements:

1. **Role + Seniority** — "Senior Backend Engineer with 8 years..."
2. **Industry/Domain Context** — "B2B SaaS in fintech, enterprise customers"
3. **Methodologies** — "Uses JTBD framework, presents at 95% confidence"
4. **Constraints** — "$50K budget, 6-week timeline, team of 3"
5. **Output Format** — "2-page executive brief with 3 options + recommendation"
Read the full skill on GitHub

Details

Skill name
prompt-improver
Path
Community/prompt-improver/SKILL.md
License
MIT
Repo updated
29 July 2026

FAQ

What does the Prompt Improver skill do?

Improve and optimize prompts for Claude using Anthropic's official best practices and the 5-Element Persona Framework. Use when asked to improve, refine, optimize, or review a prompt. Also triggers on requests to make prompts clearer, more effective, better structured, or to get better/more realistic outputs from Claude. Covers Claude 4.x-specific techniques, persona engineering, constraint-based prompting, and structured outputs.

How do I install the Prompt Improver skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add zocomputer/skills --skill prompt-improver. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Prompt Improver popular?

It lives in zocomputer/skills, which has 46 GitHub stars and 56 forks.

Related skills

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.