A curated collection of 20 specialized AI agent definitions designed to work with GitHub Copilot and similar AI coding assistants. Each agent is an expert in a specific domain, providing focused knowledge, best practices, and implementation guidance.
This repository contains pre-configured AI agents that act as specialized consultants for various software development tasks. From backend architecture to security auditing, each agent brings domain-specific expertise to your development workflow.
- backend-typescript-architect - Bun runtime, TypeScript backend architecture, database schema design, and scalable server patterns
- python-backend-engineer - FastAPI, Django development, async programming, API development, and database integration
- api-designer - Clean RESTful APIs, OpenAPI specifications, endpoint documentation, and API versioning strategies
- ui-engineer - Modern JavaScript/TypeScript frontend development, responsive design systems, React/Vue/Angular components, and performance optimization
- interface-designer - UI/UX design principles, API specification design, user experience optimization, and design system creation
- ai-engineer - LLM applications, RAG systems, prompt pipelines, and vector database management
- ml-engineer - MLOps pipelines, model serving, feature engineering, model monitoring, and production ML systems
- data-scientist - Statistical analysis, data visualization, SQL optimization, BigQuery operations, and data insights
- terraform-specialist - Infrastructure as Code design, Terraform module development, state management, and multi-cloud deployment strategies
- network-engineer - Network troubleshooting, load balancer configuration, traffic analysis, and security group management
- incident-responder - Production incidents, triage, root cause analysis, recovery procedures, and post-incident documentation
- system-architect - Comprehensive system architecture design, technology stack selection, scalability planning, and architectural documentation
- code-reviewer - Comprehensive code quality assessments, security vulnerability detection, performance analysis, and best practices enforcement
- code-refactoring-specialist - Safe code refactoring, technical debt reduction, code structure improvement, and design pattern implementation
- performance-optimizer - Analyzing and optimizing code performance, identifying bottlenecks, resource usage analysis, and scalability improvements
- test-suite-generator - Generating comprehensive test coverage, test case generation, coverage analysis, and test automation setup
- security-auditor - Vulnerability assessment, security framework analysis, compliance checking, and penetration testing guidance
- fintech-specialist - Financial models, backtesting trading strategies, market data analysis, and regulatory compliance (SOX, GDPR)
- payment-integration - Integrating Stripe, PayPal, and payment processors with PCI DSS compliance, webhook handling, and multi-currency support
- gemini-researcher - External information research, up-to-date data gathering, factual queries, and concept explanations
- Copy the agent definition file you need to your project's
.github/agents/directory - GitHub Copilot will automatically recognize and use the agent based on your context
- Reference the agent in your prompts when you need specialized assistance
Each agent file contains:
- Name: The agent's identifier
- Description: What the agent specializes in
- Tools: Available tools the agent can use
- Detailed guidance: Domain-specific best practices, patterns, and examples
Simply read through the agent files to understand their capabilities and use them as reference documentation or system prompts for your AI assistant.
Each agent definition follows this format:
---
name: agent-name
description: Brief description of the agent's capabilities
tools: [list, of, available, tools]
---
# Detailed implementation guidance
# Best practices
# Code examples
# Domain-specific patternsContributions are welcome! If you have ideas for new agents or improvements to existing ones:
- Fork the repository
- Create a new agent file following the naming convention:
agent-name.agent.md - Follow the existing structure and format
- Include comprehensive domain knowledge and practical examples
- Submit a pull request
- Focus on a specific, well-defined domain
- Include practical code examples
- Provide clear best practices
- List relevant tools and frameworks
- Add common patterns and anti-patterns
- Keep the content actionable and implementation-focused
This project is open source and available for use in your projects. Please check individual files for any specific licensing information.
- Starting a new project: Use system-architect and relevant tech-specific agents
- Code review: Use code-reviewer and security-auditor
- Performance issues: Use performance-optimizer and backend-specific agents
- Learning: Read agent files as comprehensive guides to best practices
- Production incidents: Use incident-responder for structured troubleshooting
- Combine multiple agents for complex tasks (e.g., system-architect + security-auditor + performance-optimizer)
- Use agents as learning resources - each contains condensed domain expertise
- Customize agents for your specific tech stack and requirements
- Keep agents updated with latest best practices in their domains
Made with ❤️ for developers who want specialized AI assistance