Contributing to Agentic Playbook

Guide Standards

Every guide in Agentic Playbook must be:

• Tested end-to-end — you ran every command yourself before submitting
• Self-contained — a reader can follow it from zero to working result
• Honest about failures — include a Troubleshooting section with real problems you hit
• Timestamped — include lastTested date so readers know if it's current

How to Submit a Guide

1. Fork the repo
2. Create a new directory under the appropriate tool category in src/app/
3. Write your guide as page.mdx with the required frontmatter
4. Open a PR with a description of what you tested and on what environment

Guide Template

Use this structure for every guide:

• Title — What you'll accomplish
• Prerequisites — What you need before starting
• Steps — Numbered, with actual commands and config
• Verification — How to confirm it worked
• Troubleshooting — Common failures and fixes

Frontmatter

Every guide page should start with the GuideHeader component:

<GuideHeader tool="ToolName" difficulty="Beginner" lastTested="2026-03" readTime="15 min" />

# Your Guide Title

Guide content here...

Difficulty Levels

| Level | Meaning | |-------|---------| | Beginner | No prior experience with the tool needed | | Intermediate | Assumes basic familiarity with the tool | | Advanced | Deep configuration, edge cases, production hardening |

Tools We Cover

Currently accepting guides for:

• OpenClaw — Multi-agent gateway systems
• Paperclip — AI company orchestration
• LangSmith Fleet — Agent observability
• Skills & Plugins — Extending agent capabilities
• Security — Production hardening across any tool