promptdriven/pdd

Prompt Driven Development (PDD): The Last Programming Language™. Prompt files are source; code is generated output.

What it solves

PDD (Prompt-Driven Development) addresses the issue of "specification drift" in AI-assisted coding, where the actual code diverges from the intended design. It shifts the primary source of truth from the implementation code (Python, TypeScript, Go, etc.) to human-authored .prompt files that define durable intent, constraints, examples, and tests.

How it works

PDD treats prompts as the source language and traditional code as a generated artifact. Developers write specifications in .prompt files, which are then "compiled" into implementation languages. The system provides a CLI and web interface to automate the development cycle, including implementing GitHub issues, generating architecture and UI tests, and repairing prompts. It integrates with various agentic CLIs (like Claude Code, Antigravity, and Codex) and uses LiteLLM to route requests to various LLM providers.

Who it’s for

Software developers who want to move beyond simple AI code completion toward a system where the high-level intent and specifications are the primary, reviewable source of truth for their codebase.

Highlights

  • Prompt-Native Workflow: Uses .prompt files as the source of truth instead of raw code.
  • Agentic CLI Commands: Specialized workflows for implementing feature requests (pdd change), fixing bugs (pdd bug $\rightarrow$ pdd fix), and splitting large development units (pdd split).
  • Multi-Provider Support: Integrates with multiple LLM providers via LiteLLM and supports both cloud and local execution modes.
  • Automated Quality Gates: Includes tools for prompt linting, contract checking, and coverage matrices to ensure prompt quality and others.

Related

  • Project
    gszhangwei/open-spdd

    A methodology and CLI tool that transforms AI coding prompts into structured design contracts using the REASONS Canvas framework to ensure precise, traceable code generation.

  • Project
    KhazP/vibe-coding-prompt-template

    A structured AI workflow and prompt template system that guides users from initial product research and PRD creation to shipping an MVP using AI coding agents.

  • Project
    microsoft/promptflow

    A suite of development tools for streamlining the end-to-end cycle of LLM-based AI applications, from prototyping and prompt engineering to evaluation and production deployment.

  • Project
    IBM/prompt-declaration-language

    A declarative YAML-based language for creating reliable, composable LLM prompts and integrating them into software systems with support for control flow and code execution.

  • Project
    microsoft/prompty

    A markdown-based file format and runtime for LLM prompts that decouples prompt engineering from application code, supporting Python, TypeScript, and VS Code.