Skip to main content
skill CX Codex

meta-prompt-engineering — Claude Code skill: prompts and setup

The meta-prompt-engineering skill turns unclear prompts into structured instructions with a defined role, task steps, constraints, and quality checks. Use it when an agent’s responses vary between runs.

no license 114 installs ★ 203 updated 2026-09-01

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add lyndonkl/claude --skill meta-prompt-engineering -a claude-code
Open on GitHub

What meta-prompt-engineering does

The skill examines a prompt for gaps, then helps define the AI’s role, the goal, and the expected response format. It breaks complicated work into steps and makes limits on length, tone, and content explicit. Supporting materials include resources/template.md and resources/methodology.md.

Prompts can include self-check criteria and safeguards for known failure patterns. The workflow then tests the prompt across several runs and uses resources/evaluators/rubric_meta_prompt_engineering.json to guide revisions. Repeated testing is part of the process; it does not promise identical responses.

What it helps with

  • turn a vague request into structured agent instructions
  • split a complex task into ordered steps
  • set response format and content limits
  • add checks for recurring prompt failures

Example prompts for the agent

  • Rewrite this prompt with a clear role, task steps, and response format.
  • Find why this prompt gives inconsistent results and suggest constraints to test.
  • Draft a multi-step prompt with criteria for checking the final answer.

How to install meta-prompt-engineering

1Skills CLI (Vercel)

npx skills add lyndonkl/claude --skill meta-prompt-engineering -a claude-code

2Manually: put the files in place

~/.claude/skills/meta-prompt-engineering/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/meta-prompt-engineering/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about meta-prompt-engineering

How do I install meta-prompt-engineering in Claude Code?

Run npx skills add lyndonkl/claude --skill meta-prompt-engineering -a claude-code. The skill file is at ~/.claude/skills/meta-prompt-engineering/SKILL.md.

What does meta-prompt-engineering do?

It helps shape an unclear prompt into instructions with a defined role, task stages, limits, and quality checks. It also guides testing and revisions based on recurring failures.

How does the skill test a prompt?

Run the prompt several times, identify repeated failure patterns, and add constraints to address them. The workflow includes resources/evaluators/rubric_meta_prompt_engineering.json for evaluation.

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set49

The whole set →
domain-research-health-science domain-research-health-science structures health science and medical evidence synthesis using PICOT framing, formal bias checklists, and GRADE certainty scoring. CX Codex +1 119installs adr-architecture adr-architecture turns an architecture choice into a written decision record with context, rejected options and consequences, then scores the draft against a rubric. CX Codex +1 118installs constraint-based-creativity constraint-based-creativity — a skill that makes the model invent inside tight limits, such as a small budget or a fixed format, so brainstorming stops circling the same answers. CX Codex +1 115installs translation-reframing-audience-shift translation-reframing-audience-shift rewrites content for a different audience: technical docs to business cases, strategy to action steps, expert text to plain language. CX Codex +1 115installs systems-thinking-leverage systems-thinking-leverage turns a tangled problem into a mapped system with ranked intervention points, built on feedback loops and Meadows' leverage ladder. CX Codex +1 114installs reviews-retros-reflection reviews-retros-reflection structures retrospectives, incident post-mortems, and periodic check-ins. Helps teams identify root causes and turn feedback into SMART action items without blame. CX Codex +1 113installs causal-inference-root-cause causal-inference-root-cause — a skill that pushes an agent through a five-step causal investigation, so the answer to why something happened comes with a mechanism and a tested alternative. CX Codex +1 110installs security-threat-model security-threat-model guides threat modeling across system components using STRIDE methodology, trust boundary maps, and defense-in-depth principles before audits or releases. CX Codex +1 109installs

Project sponsors

Companies that keep this analytics open