中文

CherryStudio Prompt Compiler

A Claude Code Skill for compiling local Markdown LLM prompts into CherryStudio-compatible Remote JSON format.

Overview

The CherryStudio Prompt Compiler is a zero-dependency tool that converts local Markdown-formatted LLM prompts into the JSON format required by CherryStudio. It uses Claude Code's native model capabilities to intelligently parse YAML frontmatter, generate semantic emojis, and batch-process entire folders of prompt files.

Features

• Zero External Dependencies: Uses only Claude Code built-in tools (Read, Glob, Write, Bash)
• Intelligent Emoji Generation: Leverages Claude's semantic understanding to match emojis to prompt descriptions
• Batch Processing: Process entire folders of Markdown files in one operation
• Complete Content Preservation: Retains YAML frontmatter and all Markdown formatting
• Error Tolerance: Continues processing even if individual files have issues

Installation

Prerequisites

• Claude Code installed
• A project folder with Markdown prompt files

Setup

1. Clone or copy this repository to your local machine
2. Ensure the skill directory structure is in place:

cherrystudio-prompt-compiler/
└── skills/
    └── compiling-prompts/
        ├── SKILL.md
        ├── reference/
        └── examples/

3. The skill will be automatically discovered by Claude Code

Quick Start

Basic Usage

Simply ask Claude Code to compile your prompts:

Compile all prompts in ./prompts/ folder

Or use the skill trigger:

/compiling-prompts ./my-prompts/

Input Format

Your Markdown files should include YAML frontmatter:

---
author: Your Name
version: 1.0
description: A helpful assistant for coding tasks
category:
  - Development
tags:
  - coding
  - programming
---

# Code Assistant

You are a helpful coding assistant...

Output

The compiler generates cherry-studio-prompts.json:

[
  {
    "id": "1",
    "name": "code-assistant",
    "description": "A helpful assistant for coding tasks",
    "emoji": "👨‍💻",
    "group": ["Development"],
    "prompt": "---\nauthor: ...\n---\n\n# Code Assistant\n..."
  }
]

Field Mapping

CherryStudio Field	Source	Notes
id	Auto-generated	Sequential: "1", "2", "3"...
name	Filename	Without .md extension
description	YAML description	Empty string if missing
emoji	AI-generated	Semantic matching
group	YAML category	["General"] if missing
prompt	Full content	YAML + Markdown body

Usage Examples

Compile a Single Folder

Compile prompts in ./my-prompts/

Specify Output Location

Compile prompts in ./prompts/ and save to ./output/prompts.json

Handle Errors Gracefully

The compiler will continue processing even if some files have issues:

Compilation complete with warnings:
- missing-yaml.md: No frontmatter found, using defaults
Successfully compiled: 5 files
Output: cherry-studio-prompts.json

Emoji Generation

The compiler uses semantic understanding to match emojis to your prompt descriptions:

Description Keywords	Emoji
Product manager, PM, business	👨‍💼
Developer, engineer, coding	👨‍💻
Writer, content, writing	✍️
Designer, creative, art	🎨
Analytics, data, metrics	📊
Assistant, helper, copilot	🤖
Chat, support, communication	💬
Teacher, education, learning	📚

Directory Structure

skills/compiling-prompts/
├── SKILL.md                    # Main skill file
├── reference/                  # Detailed documentation
│   ├── cherry-studio-schema.md # JSON Schema specification
│   ├── field-mapping.md        # Field mapping rules
│   ├── emoji-generation.md     # Emoji generation strategy
│   └── error-handling.md       # Error handling guide
└── examples/                   # Example files
    ├── basic-prompt.md         # Simple prompt example
    ├── advanced-prompt.md      # Complex prompt example
    └── example-output.json     # Expected output format

Importing to CherryStudio

1. Run the compiler to generate cherry-studio-prompts.json
2. Open CherryStudio
3. Go to Settings → Prompts
4. Click "Import" and select the JSON file
5. Your prompts will be available in CherryStudio

Advanced Usage

Custom Category Handling

If your prompts use non-standard category formats:

# String format (will be converted to array)
category: Template

# Array format (used as-is)
category:
  - Template
  - Product

Handling Missing Metadata

Files without YAML frontmatter will use sensible defaults:

• description: Empty string
• group: ["General"]
• emoji: Based on filename or content analysis

Troubleshooting

Problem: No Markdown files found

• Solution: Verify the directory path contains .md files

Problem: Some prompts are missing emojis

• Solution: Add a description field to your YAML frontmatter

Problem: Category is not being recognized

• Solution: Ensure category field exists in YAML frontmatter

Problem: JSON output is empty

• Solution: Check that source files are valid Markdown with content

Technical Specifications

For detailed technical information, see:

• Product Requirements Document
• Technical Specification

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

License

MIT License - See LICENSE file for details

Author

Created by Galen Dai

Version

1.1.0