# CC-Meta (Prompt Evaluator)

> Use this tool when you need to evaluate and refine AI prompts for clarity, specificity, and effectiveness, and want detailed feedback and suggestions for improvement. CC-Meta analyzes prompts using OpenAI or Anthropic models, providing numerical scores, strengths analysis, and rewrite recommendations through a convenient slash command interface. It is ideal for developers iterating on Claude Code prompts, supporting multiple AI models and customizable evaluation criteria.

Canonical page: https://skillsregistry.net/skills/cc-meta-prompt-evaluator  
JSON: https://api.skillsregistry.net/v1/skills/cc-meta-prompt-evaluator

## Description

CC-Meta (Claude Code Metaprompter) provides AI-powered prompt evaluation capabilities through an MCP server that analyzes prompts for clarity, specificity, and effectiveness using OpenAI or Anthropic models. Built with TypeScript and the Vercel AI SDK, it offers two core tools: a ping function for connection testing and an evaluate function that provides detailed feedback including numerical scores, strengths analysis, improvement suggestions, and rewrite recommendations. The implementation supports multiple AI models (OpenAI's o3, Anthropic's Claude Opus-4 and Sonnet-4) with flexible API key configuration, includes a convenient slash command interface (/meta), and features customizable evaluation criteria stored in a separate prompt template file, making it ideal for developers who want to iterate on their Claude Code prompts without leaving their terminal workflow.

## Trust

- **Trust score (0–1):** 0.97
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** ai-ml
- **Updated:** 2026-09-28

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/cc-meta-prompt-evaluator)
- **Repository:** <https://github.com/areznik23/cc-meta>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "cc-meta-prompt-evaluator"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/cc-meta-prompt-evaluator` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/cc-meta-prompt-evaluator/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt