claude-code-playbook (RiyaParikh0112) — Summary

claude-code-playbook by RiyaParikh0112 is a single-author guidebook structured as a directory of markdown documentation covering Claude Code patterns, best practices, and templates for developers. The repo contains ~20 markdown files organized across docs/ (getting-started, fundamentals, workflows, advanced, troubleshooting) plus a setup.md and no runnable primitives (zero commands, skills, agents, hooks, or MCP servers). It is a pure "Archetype 4 — Markdown scaffold" framework: the value is entirely in the template content and written guidance, not automation. The paradigm shift it teaches is from "hands-on coder" to "development director" orchestrating AI agents. Key docs include CLAUDE.md templates, slash command patterns, hooks configuration examples, multi-agent workflow descriptions, and model selection strategy (Sonnet 4 for daily work, Opus 4 for complex architecture). The stated install path is manual copy-paste of templates from the templates/ and docs/ directories into a user's project.

Differs from seeds: Closest to agent-os (Archetype 4), but entirely documentation-only — agent-os at least ships 5 runnable commands that write the markdown files; RiyaParikh0112 ships only the documentation itself. No hooks, no skills, no CLI. Unlike superpowers or bmad-method, there is no behavioral enforcement mechanism; adherence depends entirely on the developer reading and applying the written guidance.

claude-code-playbook (RiyaParikh0112) — Overview

Origin

A community playbook authored by RiyaParikh0112, framed as a "battle-tested in production environments" guide. 26 GitHub stars. No license file detected.

Core philosophy

"Transform from coder to development director: Master Claude Code to build in days what used to take months" "Claude Code isn't just AI autocomplete — it's a fundamental shift from hands-on coder to development director."

Five key principles stated:

1. Context is King — CLAUDE.md as Claude's persistent memory
2. Plan Before Code — Shift-Tab twice for architectural discussion
3. Constraint-Driven Development — describe problems + constraints, not just desired output
4. Quality Gates Are Mandatory — treat AI output as junior developer output
5. Strategic Model Selection — Sonnet 4 for daily work, Opus 4 for architecture

Critical warnings (verbatim)

• "Never use --dangerously-skip-permissions except in isolated sandboxes"
• "Always review AI-generated code before integration"
• "Use git worktrees for parallel development, not shortcuts"

Audience

Developers at any level wanting to understand Claude Code patterns; specifically mentions teams and enterprise patterns in the advanced section.

Scope note

The README lists a templates/ directory with "Ready-to-use prompts, configurations, and workflows" but at time of analysis the templates/ directory structure was not populated with files accessible via GitHub API — the content is implied by the README structure.

claude-code-playbook (RiyaParikh0112) — Architecture

Distribution type

Pure documentation repository — methodology-doc. No install commands, no packages, no binaries.

Install

Manual: read docs, copy templates to your project.

Directory structure

.
├── README.md
├── setup.md
└── docs/
    ├── getting-started/
    │   ├── installation-setup.md
    │   ├── first-session.md
    │   ├── project-initialization.md
    │   └── mindset-shift.md
    ├── fundamentals/
    │   ├── claude-md-mastery.md
    │   ├── context-management.md
    │   ├── prompt-engineering.md
    │   ├── model-selection-strategy.md
    │   └── security-best-practices.md
    ├── workflows/
    │   ├── tdd-with-claude.md
    │   ├── code-review-process.md
    │   ├── debugging-strategies.md
    │   ├── refactoring-patterns.md
    │   └── multi-agent-workflows.md
    ├── advanced/
    │   ├── hooks-and-automation.md
    │   ├── custom-configurations.md
    │   ├── ci-cd-integration.md
    │   └── scaling-strategies.md
    └── troubleshooting/
        ├── common-issues.md
        ├── performance-optimization.md
        └── debugging-guide.md

Runtime requirements

None. Pure markdown documentation.

Target AI tools

Claude Code (primary), but the documentation is tool-agnostic in content.

claude-code-playbook (RiyaParikh0112) — Components

Commands

0 — no slash commands, no bash scripts.

Skills

0 — no skill files.

Agents / Subagents

0 — no agent definitions. The docs/workflows/multi-agent-workflows.md discusses multi-agent patterns conceptually but ships no agent files.

Hooks

0 — no settings.json, no hook scripts. docs/advanced/hooks-and-automation.md provides example hook configuration snippets (shown as code blocks in documentation) but these are illustrative, not deployed.

MCP servers

Templates (implied, not confirmed populated)

README mentions:

• CLAUDE.md Templates
• Slash Commands (custom workflow automation)
• Hooks (quality gate automation)
• Prompts (proven prompt patterns)

Documentation modules (20 markdown files)

claude-code-playbook (RiyaParikh0112) — Uniqueness

Differs from seeds

Closest to agent-os (Archetype 4), but even more minimal: agent-os ships 5 runnable commands that write the markdown; this playbook ships only the documentation itself with zero runnable primitives. The hook examples in docs/advanced/hooks-and-automation.md use event names (preEdit, postEdit) that don't match the current Claude Code API, suggesting some content was written before or without testing against the live tool. Unlike bmad-method (34 skills, 6 subagent personas) or superpowers (14 skills, behavioral enforcement), this is purely advisory. It is most similar to a comprehensive README-as-methodology document.

Positioning

Community-authored "getting started" guide for Claude Code. Strong on the CLAUDE.md/context-management/model-selection conceptual side. Weak on automation. Targets developers who want to understand the paradigm before applying any tooling.

Observable failure modes

1. Hook event names may be outdated — preEdit/postEdit format in docs differs from actual Claude Code API (PreToolUse/PostToolUse).
2. No enforcement mechanism — all guidance is advisory; no hooks, skills, or agents enforce the described patterns.
3. Template directory may be empty — the README promises templates but the directory structure shows only top-level docs.
4. Single-author, low-star project — 26 stars, last commit 2025-09-23 (inactive); maintenance unclear.