shinpr/claude-code-workflows

Development workflows for Claude Code that keep broad exploration focused on the outcome you approved.

What it solves

Claude Code can sometimes lose focus during complex codebase exploration, spending too much time on side-findings rather than the intended goal. This project provides a structured set of workflows (as plugins) to ensure that AI-driven development stays pointed at an agreed-upon result, preventing scope creep and ensuring that implementation details are verified against design decisions.

How it works

It operates as a series of "recipes" (plugins) for Claude Code that guide the AI through a rigorous development lifecycle. The process typically follows these steps:

  1. Agreement: The AI and user agree on the desired outcome and exclusions before any code is written.
  2. Design: For non-trivial changes, the AI creates Design Docs, PRDs, or UI Specs that are reviewed and approved.
  3. Implementation: The AI breaks the work into a Work Plan with specific tasks. Each task is implemented, verified, and quality-checked before being committed.
  4. Verification: An independent review is performed on the finished implementation to ensure it meets the agreed outcome and contains no security or reliability issues.

Who it’s for

Developers using Claude Code who are working on non-trivial changes that require scope agreement, durable design decisions, and reliable verification, particularly for backend, frontend (React/TypeScript), or full-stack projects.

Highlights

  • Specialized Recipes: Dedicated paths for backend, frontend, and full-stack development, including specific tools for UI analysis and React Testing Library checks.
  • Independent Review: Uses separate agent calls to perform security and functional reviews to avoid the "self-grading" bias of a single session.
  • Lite Mode: A streamlined version of the workflow that reduces the number of checks for faster iteration while maintaining approval stops.
  • Comprehensive Tooling: Includes recipes for diagnosing problems, reverse-engineering existing systems into documentation, and adjusting UIs against design sources.

Related

  • Project
    umputun/cc-thingz

    A marketplace of Claude Code plugins that add brainstorming, PR review, planning, release automation, thinking aids, skill‑evaluation hooks, and workflow helpers. Install via `/plugin marketplace add umputun/cc-thingz` and pick the plugins you need.

  • Project
    pedrohcgs/claude-code-my-workflow

    A ready-to-fork foundation for AI-assisted academic work that uses Claude Code to automate the planning, creation, and verification of research papers, slides, and data analysis.

  • Project
    Chachamaru127/claude-code-harness

    A management framework for AI coding assistants that adds structured planning, safety guardrails, and mandatory review steps to ensure reliable code generation.

  • Dispatch
    Claude Code Best Practices Guide

    Anthropic provides a comprehensive set of best practices for optimizing Claude Code, an agentic coding environment, focusing on context window management and autonomous verification.