Claude Agents

A comprehensive collection of AI agent definitions for Claude Code, designed to simulate a full-stack startup team.

Overview

This repository contains 34 specialized agent prompts organized into 7 categories, covering the complete product lifecycle from ideation to operation.

.claude/agents/
├── engineering/        # 6 agents - Build the product
├── product/            # 3 agents - Define what to build
├── marketing/          # 7 agents - Grow the user base
├── design/             # 5 agents - Create the experience
├── project-management/ # 3 agents - Ship on time
├── studio-operations/  # 5 agents - Run the business
└── testing/            # 5 agents - Ensure quality

Agent Categories

Engineering (6 agents)

Agent	Description
frontend-developer	React/Vue development, component architecture, performance optimization
backend-architect	System design, API development, database architecture, scalability
mobile-app-builder	React Native/Flutter cross-platform and native iOS/Android development
ai-engineer	LLM integration, RAG systems, agent development, prompt engineering
devops-automator	CI/CD pipelines, containerization, Kubernetes, monitoring
rapid-prototyper	MVP development, quick validation, speed over perfection

Product (3 agents)

Agent	Description
trend-researcher	Market intelligence, competitor analysis, emerging opportunities
feedback-synthesizer	User feedback analysis, pattern recognition, actionable insights
sprint-prioritizer	Backlog prioritization using RICE/MoSCoW, sprint planning

Marketing (7 agents)

Agent	Description
tiktok-strategist	Short-form video strategy, viral content, platform growth
instagram-curator	Visual content curation, Feed/Stories/Reels planning
twitter-engager	Tweet/thread writing, community engagement, thought leadership
reddit-community-builder	Authentic community presence, organic engagement
app-store-optimizer	ASO for iOS/Android, keyword optimization, conversion
content-creator	Blog posts, video scripts, newsletters, multi-platform distribution
growth-hacker	Growth experiments, funnel optimization, AARRR metrics

Design (5 agents)

Agent	Description
ui-designer	Design systems, component design, spacing/typography/color
ux-researcher	User interviews, usability testing, research insights
brand-guardian	Brand consistency, guidelines enforcement, quality control
visual-storyteller	Visual narratives, infographics, presentation design
whimsy-injector	Micro-interactions, easter eggs, delightful moments

Project Management (3 agents)

Agent	Description
experiment-tracker	A/B test management, hypothesis tracking, learnings capture
project-shipper	Release management, launch checklists, risk assessment
studio-producer	Creative project coordination, resource allocation, timeline management

Studio Operations (5 agents)

Agent	Description
support-responder	Customer support with HEART method, empathetic responses
analytics-reporter	Data analysis, weekly/monthly reports, actionable insights
infrastructure-maintainer	System monitoring, incident response, maintenance runbooks
legal-compliance-checker	GDPR/CCPA compliance, privacy policy review, risk assessment
finance-tracker	MRR/ARR tracking, cash flow, budget management, unit economics

Testing (5 agents)

Agent	Description
tool-evaluator	Software evaluation, vendor assessment, adoption recommendations
api-tester	API testing, functional/security/performance validation
workflow-optimizer	Process analysis, waste identification, efficiency improvements
performance-benchmarker	Load testing, performance baselines, bottleneck identification
test-results-analyzer	Test result interpretation, failure patterns, quality gates

Usage

With Claude Code

These agents are designed to work with Claude Code. Place the .claude/agents/ directory in your project root, and Claude Code will have access to these agent definitions.

Agent Structure

Each agent file follows a consistent structure:

# Agent Name

[Role description and philosophy]

## Core Expertise / Responsibilities
[What this agent specializes in]

## Frameworks / Methodologies
[Industry-standard approaches used]

## Templates / Checklists
[Ready-to-use templates]

## Best Practices
[Guidelines and recommendations]

## Output Format
[Expected deliverable structure]

Key Features

• Role Clarity: Each agent has well-defined responsibilities and boundaries
• Plug & Play: Includes frameworks, templates, and checklists ready for immediate use
• Standardized Output: Consistent output formats for easy collaboration
• Best Practices Built-in: Industry methodologies (RICE, HEART, AARRR, etc.) embedded

Contributing

Contributions are welcome! When adding new agents:

1. Follow the existing file structure and naming conventions
2. Include clear role definition and responsibilities
3. Add practical templates and checklists
4. Define expected output formats

License

MIT License - feel free to use and modify for your projects.