#279279th of 455 in Skills, plugins and prompts
pdd
Prompt-first development CLI and agent skills with simulation before code is written
- Stars
- 935
- License
- MIT
- Last commit
- Nov 2026
- Last release
- Nov 2026
- Language
- JavaScript
631 on 22 Apr 2026935 on 11 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
PDD treats prompts and user Stories as the program source. A hosted interpreter runs tests and simulations (GUI, CLI or HTTP API) from the prompts alone, and after approval your agent generates code and reruns the same tests. It ships as the `pdd` Node CLI plus skills for Codex, Claude Code, Gemini CLI and Antigravity CLI, with an optional Autopilot mode that works GitHub issues into PRs.
Who it is for: Developers who drive coding agents and want to approve behavior before code
Strengths
- Simulate a screen, CLI or HTTP API from prompts before any code exists
- Per-run AI-call budgets: 400 per simulation, 12 per test step (min 18, max 1000)
- Works with Codex, Claude Code, Gemini CLI and Antigravity CLI via installable skills
- Approval gates keep a human sign-off before code generation; MIT licensed
Weaknesses
- Test and simulate need the hosted interpreter and an approved PDD Cloud GitHub account
- Workflow runs give agents full access and need Herdr installed
- Tested only on macOS and Linux; requires Node.js 22.16+ and an authenticated gh CLI
- Code-space evidence is agent-reported, not independently verified execution
What it needs
- no GPU
- Needs Node.js 22.16+, Git, GitHub CLI, Codex, Claude Code, Gemini CLI or Antigravity CLI, Herdr (workflow runs and Autopilot), Hosted PDD interpreter (PDD Cloud)
- Models: Codex, Claude Code, Gemini CLI, Antigravity CLI
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in Skills, plugins and prompts
See all 455| Rank | Project | Score |
|---|---|---|
| 1 | claude-memPersistent cross-session memory plugin for Claude Code and other coding agents | 97 out of 100 |
| 2 | impeccableDesign skill with 24 slash commands for AI coding agents' frontend output | 97 out of 100 |
| 3 | gstackMarkdown slash-command skills that give Claude Code planning, review, QA and ship roles | 96 out of 100 |
| #4 | rufloAgent harness adding swarms, memory and hooks to Claude Code and Codex | 96 out of 100 |
| #5 | cavemanShortens coding-agent replies and compresses tool output before the model reads it | 95 out of 100 |