[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-72561":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":19,"compositeScore":20,"rankGlobal":9,"rankLanguage":9,"license":9,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":25,"readmeContent":26,"aiSummary":27,"trendingCount":15,"starSnapshotCount":15,"syncStatus":28,"lastSyncTime":29,"discoverSource":30},72561,"PRPs-agentic-eng","Wirasm\u002FPRPs-agentic-eng","Wirasm","Prompts, workflows and more for agentic engineering ",null,"Python",2192,610,70,8,0,6,12,30,18,78.36,false,"development",true,[],"2026-06-12 04:01:06","# PRP (Product Requirement Prompts)\n\nA collection of prompts for AI-assisted development with Claude Code.\n\n## Video Walkthrough\n\nhttps:\u002F\u002Fwww.youtube.com\u002Fwatch?v=KVOZ9s1S9Gk&lc=UgzfwxvFjo6pKEyPo1R4AaABAg\n\n### Support This Work\n\n**Found value in these resources?**\n\n**Buy me a coffee:** https:\u002F\u002Fcoff.ee\u002Fwirasm\n\nI spent a considerable amount of time creating these resources and prompts. If you find value in this project, please consider buying me a coffee to support my work.\n\n---\n\n### Transform Your Team with AI Engineering Workshops\n\n**Ready to move beyond toy demos to production-ready AI systems?**\n\n**Book a workshop:** https:\u002F\u002Fwww.rasmuswiding.com\u002F\n\n**What you'll get:**\n\n- Put your team on a path to become AI power users\n- Learn the exact PRP methodology used by top engineering teams\n- Hands-on training with Claude Code, PRPs, and real codebases\n- From beginner to advanced AI engineering workshops for teams and individuals\n\n**Perfect for:** Engineering teams, Product teams, and developers who want AI that actually works in production\n\nContact me directly at hello@rasmuswiding.com\n\n---\n\n## What is PRP?\n\n**Product Requirement Prompt (PRP)** = PRD + curated codebase intelligence + agent\u002Frunbook\n\nThe minimum viable packet an AI needs to ship production-ready code on the first pass.\n\nA PRP supplies an AI coding agent with everything it needs to deliver a vertical slice of working software—no more, no less.\n\n### How PRP Differs from Traditional PRD\n\nA traditional PRD clarifies _what_ the product must do and _why_ customers need it, but deliberately avoids _how_ it will be built.\n\nA PRP keeps the goal and justification sections of a PRD yet adds AI-critical layers:\n\n- **Context**: Precise file paths, library versions, code snippet examples\n- **Patterns**: Existing codebase conventions to follow\n- **Validation**: Executable commands the AI can run to verify its work\n\n---\n\n## Quick Start\n\n### Option 1: Copy Commands to Your Project\n\n```bash\n# From your project root\ncp -r \u002Fpath\u002Fto\u002FPRPs-agentic-eng\u002F.claude\u002Fcommands\u002Fprp-core .claude\u002Fcommands\u002F\n```\n\n### Option 2: Clone Repository\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FWirasm\u002FPRPs-agentic-eng.git\ncd PRPs-agentic-eng\n```\n\n---\n\n## Commands\n\nThe `.claude\u002Fcommands\u002Fprp-core\u002F` directory contains the core PRP workflow commands:\n\n### Core Workflow\n\n| Command          | Description                                              |\n| ---------------- | -------------------------------------------------------- |\n| `\u002Fprp-prd`       | Interactive PRD generator with implementation phases     |\n| `\u002Fprp-plan`      | Create implementation plan (from PRD or free-form input) |\n| `\u002Fprp-implement` | Execute a plan with validation loops                     |\n\n### Issue & Debug Workflow\n\n| Command                  | Description                                      |\n| ------------------------ | ------------------------------------------------ |\n| `\u002Fprp-issue-investigate` | Analyze GitHub issue, create implementation plan |\n| `\u002Fprp-issue-fix`         | Execute fix from investigation artifact          |\n| `\u002Fprp-debug`             | Deep root cause analysis with 5 Whys methodology |\n\n### Git & Review\n\n| Command       | Description                                       |\n| ------------- | ------------------------------------------------- |\n| `\u002Fprp-commit` | Smart commit with natural language file targeting |\n| `\u002Fprp-pr`     | Create PR with template support                   |\n| `\u002Fprp-review` | Comprehensive PR code review                      |\n\n### Autonomous Loop\n\n| Command             | Description                                      |\n| ------------------- | ------------------------------------------------ |\n| `\u002Fprp-ralph`        | Start autonomous loop until all validations pass |\n| `\u002Fprp-ralph-cancel` | Cancel active Ralph loop                         |\n\n---\n\n## Ralph Loop (Autonomous Execution)\n\nBased on [Geoffrey Huntley's Ralph Wiggum technique](https:\u002F\u002Fghuntley.com\u002Fralph\u002F) - a self-referential loop that keeps iterating until the job is actually done.\n\n### How It Works\n\n```\n\u002Fprp-ralph .claude\u002FPRPs\u002Fplans\u002Fmy-feature.plan.md --max-iterations 20\n```\n\n1. Claude implements the plan tasks\n2. Runs all validation commands (type-check, lint, tests, build)\n3. If any validation fails → fixes and re-validates\n4. Loop continues until ALL validations pass\n5. Outputs `\u003Cpromise>COMPLETE\u003C\u002Fpromise>` and exits\n\nEach iteration, Claude sees its previous work in files and git history. It's not starting fresh - it's debugging itself.\n\n### Setup\n\nThe stop hook must be configured in `.claude\u002Fsettings.local.json`:\n\n```json\n{\n  \"hooks\": {\n    \"Stop\": [\n      {\n        \"hooks\": [\n          {\n            \"type\": \"command\",\n            \"command\": \".claude\u002Fhooks\u002Fprp-ralph-stop.sh\"\n          }\n        ]\n      }\n    ]\n  }\n}\n```\n\n### Usage\n\n```bash\n# Create a plan\n\u002Fprp-plan \"add user authentication with JWT\"\n\n# Let Ralph loose\n\u002Fprp-ralph .claude\u002FPRPs\u002Fplans\u002Fadd-user-auth.plan.md --max-iterations 20\n\n# Cancel if needed\n\u002Fprp-ralph-cancel\n```\n\n### Tips\n\n- Always use `--max-iterations` (default: 20) to prevent infinite loops\n- Works best with plans that have clear, testable validation commands\n- State is tracked in `.claude\u002Fprp-ralph.state.md`\n- Progress and learnings are captured in the implementation report\n\n---\n\n## Workflow Overview\n\n### Large Features: PRD → Plan → Implement\n\n```\n\u002Fprp-prd \"user authentication system\"\n    ↓\nCreates PRD with Implementation Phases table\n    ↓\n\u002Fprp-plan .claude\u002FPRPs\u002Fprds\u002Fuser-auth.prd.md\n    ↓\nAuto-selects next pending phase, creates plan\n    ↓\n\u002Fprp-implement .claude\u002FPRPs\u002Fplans\u002Fuser-auth-phase-1.plan.md\n    ↓\nExecutes plan, updates PRD progress, archives plan\n    ↓\nRepeat \u002Fprp-plan for next phase\n```\n\n### Medium Features: Direct to Plan\n\n```\n\u002Fprp-plan \"add pagination to the API\"\n    ↓\nCreates implementation plan from description\n    ↓\n\u002Fprp-implement .claude\u002FPRPs\u002Fplans\u002Fadd-pagination.plan.md\n```\n\n### Bug Fixes: Issue Workflow\n\n```\n\u002Fprp-issue-investigate 123\n    ↓\nAnalyzes issue, creates investigation artifact\n    ↓\n\u002Fprp-issue-fix 123\n    ↓\nImplements fix, creates PR\n```\n\n---\n\n## Artifacts Structure\n\nAll artifacts are stored in `.claude\u002FPRPs\u002F`:\n\n```\n.claude\u002FPRPs\u002F\n├── prds\u002F              # Product requirement documents\n├── plans\u002F             # Implementation plans\n│   └── completed\u002F     # Archived completed plans\n├── reports\u002F           # Implementation reports\n├── issues\u002F            # Issue investigation artifacts\n│   └── completed\u002F     # Archived completed investigations\n└── reviews\u002F           # PR review reports\n```\n\n---\n\n## PRD Phases\n\nPRDs include an Implementation Phases table for tracking progress:\n\n```markdown\n| #   | Phase | Description | Status      | Parallel | Depends | PRP Plan |\n| --- | ----- | ----------- | ----------- | -------- | ------- | -------- |\n| 1   | Auth  | User login  | complete    | -        | -       | [link]   |\n| 2   | API   | Endpoints   | in-progress | -        | 1       | [link]   |\n| 3   | UI    | Frontend    | pending     | with 4   | 2       | -        |\n| 4   | Tests | Test suite  | pending     | with 3   | 2       | -        |\n```\n\n- **Status**: `pending` → `in-progress` → `complete`\n- **Parallel**: Phases that can run concurrently (in separate worktrees)\n- **Depends**: Phases that must complete first\n\n---\n\n## PRP Best Practices\n\n1. **Context is King**: Include ALL necessary documentation, examples, and caveats\n2. **Validation Loops**: Provide executable tests\u002Flints the AI can run and fix\n3. **Information Dense**: Use keywords and patterns from the codebase\n4. **Bounded Scope**: Each plan should be completable by an AI in one loop\n\n---\n\n## Project Structure\n\n```\nyour-project\u002F\n├── .claude\u002F\n│   ├── commands\u002Fprp-core\u002F   # PRP commands\n│   ├── PRPs\u002F                # Generated artifacts\n│   └── agents\u002F              # Custom subagents\n├── PRPs\u002F\n│   ├── templates\u002F           # PRP templates\n│   └── ai_docs\u002F             # Library documentation\n├── CLAUDE.md                # Project-specific guidelines\n└── src\u002F                     # Your source code\n```\n\n---\n\n## Parallel Development with Worktrees\n\nWhen PRD phases can run in parallel:\n\n```bash\n# Phase 3 and 4 can run concurrently\ngit worktree add -b phase-3-ui ..\u002Fproject-phase-3\ngit worktree add -b phase-4-tests ..\u002Fproject-phase-4\n\n# Run Claude in each\ncd ..\u002Fproject-phase-3 && claude\ncd ..\u002Fproject-phase-4 && claude\n```\n\n---\n\n## Resources\n\n### Templates (PRPs\u002Ftemplates\u002F)\n\n- `prp_base.md` - Comprehensive PRP template\n- `prp_story_task.md` - Story\u002Ftask template\n- `prp_planning.md` - Planning template\n\n### AI Documentation (PRPs\u002Fai_docs\u002F)\n\nCurated documentation for Claude Code context injection.\n\n### Legacy Commands\n\nPrevious command versions are preserved in `old-prp-commands\u002F` for reference.\n\n---\n\n## License\n\nMIT License\n\n---\n\n## Support\n\nI spent a considerable amount of time creating these resources and prompts. If you find value in this project, please consider buying me a coffee to support my work.\n\n**Buy me a coffee:** https:\u002F\u002Fcoff.ee\u002Fwirasm\n\n---\n\n**The goal is one-pass implementation success through comprehensive context.**\n","PRPs-agentic-eng 是一个用于辅助AI开发的项目，特别是针对使用Claude Code进行编程的情况。该项目提供了一系列产品需求提示（PRP），这些提示结合了产品需求文档、精选代码库智能和执行指南，旨在帮助AI一次性交付生产就绪的代码。其核心功能包括交互式生成产品需求文档、创建实施计划以及通过验证循环执行计划等，所有这些都是为了确保AI能够理解并遵循现有代码库的规范。适合于希望将AI技术应用于实际软件开发中的工程团队、产品经理及开发者使用，尤其对于那些寻求从基础到高级AI工程培训的团队来说非常有价值。",2,"2026-06-11 03:42:34","high_star"]