Claude Code Workflows
Claude Code Workflows
Section titled “Claude Code Workflows”Step-by-step guides for common development patterns with Claude Code.
🔍 Search & Discovery
Section titled “🔍 Search & Discovery”Master the art of code search by combining rg, grepai, Serena & ast-grep
Learn when to use each tool, how to combine them for maximum efficiency, and real-world workflows including:
- Exploring unknown codebases
- Large-scale refactoring
- Security audits
- Framework migrations
- Performance optimization
Key Topics:
- Quick decision matrix
- Complete feature comparison
- 5 combined workflows
- Performance benchmarks
- Common pitfalls
- Tool selection cheatsheet
🎯 Development Workflows
Section titled “🎯 Development Workflows”Structure complex tasks with planning mode before execution.
When to use: Multi-step features, architectural changes, uncertainty about approach
Test-Driven Development workflow: write tests first, implement after.
When to use: Critical functionality, regression prevention, API design
Write specifications before code for better requirements clarity.
When to use: Team collaboration, complex features, documentation-first projects
Improve code through multiple refinement cycles.
When to use: Quality improvements, performance optimization, code cleanup
Use existing battle-tested repositories as scaffolding for new projects.
When to use: Starting new projects, standardizing team patterns, rapid prototyping from proven foundations
Scale CLAUDE.md across a multi-developer, multi-tool team with profile-based module assembly.
When to use: Team 5+ devs, multiple AI tools (Claude Code + Cursor/Windsurf), mixed OS
Enforce per-PR documentation with a 3-layer system: CLAUDE.md rule + UserPromptSubmit hook + CI gate
Eliminates merge conflicts on CHANGELOG.md, captures context at implementation time, and ensures DB migrations are never silently deployed. Includes a reusable UserPromptSubmit hook pattern for enforcing any mandatory workflow step.
Key Topics:
- CLAUDE.md workflow rule for autonomous fragment creation
UserPromptSubmithook with 3-tier priority (enforcement, discovery, contextual)- Conditional suggestion pattern: “if PR-intent without fragment-mention”
- CI enforcement with independent migration check job
Two-engine UserPromptSubmit hook system: regex enforcement + self-calibrating BM25 lexical scoring
Injects ranked skill suggestions as additionalContext so Claude proposes the right skill even when phrasing was never anticipated. BM25 generalizes from scenario examples; the regex layer handles enforcement patterns. Both run in parallel on every prompt.
Key Topics:
- Regex vs BM25 decision table (when each engine wins)
- Corpus format: 10+ positive + 3+ negative scenarios per skill
build-index.js: leave-one-out calibration, F-beta threshold,ok/excluded/conflictstatus- Wiring two hooks in parallel via settings.json
- Detached background index rebuild when corpus changes
3-phase feature development with explicit validation gates between phases
Build features in three locked phases: Research feasibility first, plan the implementation second, write code third. Each phase produces a concrete artifact (RESEARCH.md → PLAN.md → code). Each gate requires an explicit GO before the next phase starts.
When to use: Features with unclear feasibility, more than a day of work, unknown technical territory, or anywhere discovering a wrong assumption late is costly
5 production-ready patterns for automating PR reviews, issue triage, and quality gates
Connect Claude directly to your GitHub workflow via the official claude-code-action. Two modes: interactive (@claude mentions) and fully automated (push/schedule triggers).
Key Topics:
- Setup via
/install-github-app(30-second quickstart) - Pattern 1: On-demand PR review via
@claudemention - Pattern 2: Automatic review on every push
- Pattern 3: Issue triage and labeling
- Pattern 4: Security-focused review on sensitive paths
- Pattern 5: Scheduled weekly repo health check
- Cost control, concurrency, fork safety
When to use: Any team wanting AI-powered code review without managing infrastructure
Run Claude Code Action alongside CodeRabbit and Greptile without triplicate findings
Non-redundant architecture for three automated reviewers on the same PR: Claude owns deep semantic review and the merge-blocking gate, a deterministic tool owns PASS/FAIL pre-merge checks, a RAG tool owns cross-file invariants. Covers the CI gate script, batching for large PRs, delta-review, and cross-tool deduplication.
When to use: Running more than one automated code reviewer and finding the same issue reported three times, or wanting Claude’s findings to actually block merge instead of just commenting
Switch between specialist roles across your ship cycle: strategic product gate, architecture review, paranoid code review, automated release, native browser QA, and retrospective.
When to use: Ship cycles where you want explicit separation between product direction, engineering rigor, review, and release — rather than one generic assistant handling all phases
🎨 Design & Content
Section titled “🎨 Design & Content”Convert design mockups (Figma, wireframes) into working code.
When to use: Frontend development, UI implementation, design system work
Generate social preview images dynamically at build time with Satori and resvg.
When to use: Astro projects, keeping social previews accurate without maintaining static PNGs
Generate professional PDFs using Quarto/Typst with Claude Code.
When to use: Reports, documentation, whitepapers, technical documents
6-stage skill pipeline: raw material → structured talk → AI-generated slides via Kimi.
When to use: Conference talks, meetup presentations, internal tech talks — from article, transcript, or notes
Configure Text-to-Speech for Claude Code responses (Agent Vibes integration).
When to use: Audio feedback, accessibility, hands-free coding
🔬 Code Exploration
Section titled “🔬 Code Exploration”Systematically explore and understand unfamiliar codebases.
When to use: New projects, legacy code, documentation gaps
Related: See Search Tools Mastery for advanced multi-tool exploration strategies.
Multi-Agent & Advanced
Section titled “Multi-Agent & Advanced”JavaScript scripts that orchestrate tens to hundreds of subagents with deterministic control flow, parallel fan-out, and automatic resume on interruption.
When to use: Multi-stage parallelizable work, long-running jobs, tasks needing structured data handoffs between phases, or any situation where a single agent would hit context limits
Key Topics:
- Primitive reference:
agent(),parallel(),pipeline(),phase(),log(),budget pipeline()vsparallel()with benchmark data (3x latency difference, 2x token difference)- Schema-structured outputs across phases
- 5 named patterns: adversarial verification, loop-until-dry, judge panels, multi-modal sweep, plan-execute-review
- Behavioral guarantees: determinism constraints, resume/caching, concurrency caps
Orchestrate multiple specialized agents working in parallel on complex tasks.
When to use: Tasks that benefit from parallelism, specialized expertise, or independent verification
Fast-track guide to setting up your first agent team in under 30 minutes.
When to use: New to multi-agent patterns, want to experiment before committing to full setup
Run Opus for planning and Sonnet for execution in two coordinated Claude Code instances.
When to use: Complex features needing deep reasoning for architecture, cost-effective execution
Coordinate agents through hook events rather than direct orchestration.
When to use: Reactive workflows, hook-triggered automation, loosely-coupled agent pipelines
Full end-to-end plan pipeline: /plan-start, /plan-validate, /plan-execute as a coherent workflow.
When to use: Any significant feature where planning rigor pays off before writing code
Multi-session task tracking with TodoWrite, tasks API, and context persistence across sessions.
When to use: Long-running tasks spanning multiple sessions, team coordination, complex backlogs
Quick Selection Guide
Section titled “Quick Selection Guide”| Your Situation | Recommended Workflow |
|---|---|
| New to codebase | Exploration Workflow + Search Tools Mastery |
| Complex feature | Plan-Driven or Spec-First |
| Need reliability | TDD with Claude |
| Large refactoring | Search Tools Mastery |
| UI implementation | Design to Code |
| Code quality | Iterative Refinement |
| New project from template | Skeleton Projects |
| Team AI instructions | Team AI Instructions |
| Enforce mandatory workflow steps | Changelog Fragments |
| Skill routing from natural-language prompts | Smart-Suggest Routing |
| Unknown feasibility, multi-day feature | RPI: Research → Plan → Implement |
| Documentation | PDF Generation |
| Social previews | OG Image Generation |
| Conference talk from raw material | Talk Preparation Pipeline |
| Audio feedback | TTS Setup |
| Deterministic multi-agent orchestration | Dynamic Workflows |
| Multi-agent tasks | Agent Teams |
| First agent team | Agent Teams Quick Start |
| Cost-optimized planning | Dual-Instance Planning |
| Hook-driven automation | Event-Driven Agents |
| Full plan workflow | Plan Pipeline |
| Multi-session tracking | Task Management |
| Strategic gate before coding | Cognitive Mode Switching |
| Non-MCP browser automation | Cognitive Mode Switching |
Contributing
Section titled “Contributing”New workflow ideas? Open an issue or PR in the main repository.
Workflow Template Structure:
- Title & Purpose
- When to Use
- Prerequisites
- Step-by-Step Guide
- Real-World Examples
- Common Pitfalls
- Related Workflows
Last updated: March 2026