Agent-Skills.md

Agent Skills: Prompt Converter

Convert between Claude Code command markdown files (*.md) and GitHub Copilot prompt files (*.prompt.md). Use when porting custom prompts or CLI commands between ecosystems.

UncategorizedID: arisng/github-copilot-fc/prompt-converter

Repository

arisng/github-copilot-fc
arisng
4

Install this agent skill to your local

pnpm dlx add-skill https://github.com/arisng/github-copilot-fc/tree/HEAD/skills/prompt-converter

Skill Files

Browse the full folder contents for prompt-converter.

Download Skill

Loading file tree…

skills/prompt-converter/SKILL.md

Skill Metadata

Name
prompt-converter
Description
'Convert between Claude Code command markdown files (*.md) and GitHub Copilot prompt files (*.prompt.md). Use when porting custom prompts or CLI commands between ecosystems.'

Prompt Converter

This skill automates the bidirectional conversion between Claude Code .md command files and Copilot .prompt.md files, ensuring metadata and instructions are correctly mapped.

Environment Detection

Before executing any operations, the skill defines the skills_source path:

import os
skills_source = os.path.expanduser('~/.copilot/skills') if os.name == 'nt' else os.path.expanduser('~/.copilot/skills')

Workflow

  1. Analyze Source: Read the source file (Claude .md or Copilot .prompt.md) to extract metadata (name, description, agent, tools) and instruction content.
  2. Map Metadata:
    • Claude to Copilot: Convert frontmatter keys, ensuring name, description, and agent (default: 'agent') are present.
    • Copilot to Claude: Convert frontmatter to Claude's expected format (often simpler description/tools).
  3. Transform Content:
    • Adjust template variables (e.g., \$ARGUMENTS in Copilot).
    • Ensure markdown structures are preserved.
  4. Execute Conversion: Use the provided script to perform the transformation. 
    python {skills_source}/prompt-converter/scripts/convert_prompt.py --src "my-command.md" --to copilot
  5. Verify: Confirm the new file exists in the target directory (e.g., prompts/ for Copilot).

Principles

  • Metadata Integrity: Preserve as much context as possible during mapping.
  • Target Specifics: Use ${input:...} for Copilot parameters and $ARGUMENTS for CLI inputs where appropriate.
  • Idempotency: Converting back and forth should result in minimal structural drift.

Usage

Convert Claude to Copilot

python scripts/convert_prompt.py --src path/to/claude.md --to copilot -o ./prompts

Convert Copilot to Claude

python scripts/convert_prompt.py --src path/to/copilot.prompt.md --to claude -o ./claude-commands

Resources

scripts/

  • convert_prompt.py: Core logic for conversion.

templates/

  • copilot.prompt.md.template: Boilerplate for Copilot prompts.
  • claude.md.template: Boilerplate for Claude commands.