mrwogu/promptscript

Prompt-as-Code for Enterprise AI. Standardize, audit, and deploy instructions across any AI coding assistant.

What it solves

PromptScript addresses the problem of "configuration drift" across different AI coding tools. Because different platforms (like Claude Code, Cursor, and GitHub Copilot) use different file formats and schemas for instructions, skills, and agent settings, managing them manually leads to inconsistencies across teams and repositories.

How it works

PromptScript acts as a "configuration as code" compiler. You define your agent platform—including instructions, skills, specialist agents, MCP servers, and workflows—using a single, validated, Git-native source language. The tool then compiles this source into the native configuration files that 50 different AI coding platforms understand, ensuring each tool receives the richest representation possible without using a runtime proxy.

Who it’s for

Developers and engineering organizations that use multiple AI coding assistants and want a single source of truth for their project's AI instructions, standards, and agent capabilities.

Highlights

  • Multi-platform support: Compiles to 50 targets, including rich native formatters for Claude Code, Cursor, and GitHub Copilot.
  • Composable architecture: Use @inherit, @use, and @extend to share and version versioned standards and skills across projects.
  • Portable skills: Define skills as standalone directories with references and scripts that can be imported from Git or registries.
  • Enterprise governance: Includes a policy engine, lockfiles for reproducible builds, and private registries for sharing standards.
  • C-I ready: Provides a CLI with strict validation and machine-readable JSON output for automation.
  • Migration tools: Ability to import existing instruction files (like CLAUDE.md) and migrate them into PromptScript source.

Related

  • Project
    alirezarezvani/claude-skills

    A comprehensive library of 388 modular agent skills and plugins for AI coding tools, providing specialized domain expertise across engineering, business, and research.

  • Project
    doccker/cc-use-exp

    A maintainable configuration system that synchronizes coding rules, skills, and workflows across multiple AI coding assistants like Claude Code, Cursor, and Gemini CLI.

  • Project
    BigSweetPotatoStudio/HyperChat

    A local AI Agent platform that implements an "AI as Code" philosophy, allowing project-specific AI assistants to be configured, versioned, and managed locally via files.

  • Project
    wshobson/agents

    A marketplace of portable agentic building blocks that provides agents, skills, and commands compatible with multiple AI coding assistants like Claude Code and Cursor.

  • Project
    EveryInc/compound-engineering-plugin

    A plugin providing 35 AI coding skills that implements a structured loop of brainstorming, planning, and reviewing to capture and compound engineering knowledge within a codebase.