- name
- agency-agents
- description
- Activate specialized AI agent personas for targeted expertise. The Agency provides pre-built personas across 18+ domains including engineering, marketing, product, sales, design, and specialized roles. Use when you need to adopt a focused professional identity with proven workflows, communication style, and domain expertise. Common triggers: "Activate Frontend Developer mode", "Switch to Product Manager", "Use Data Analyst persona", or when working on domain-specific tasks that benefit from specialized perspective and proven best practices.
Agency Agents
The Agency is a collection of meticulously crafted AI agent personalities, each with:
- Deep expertise in their domain (not generic templates)
- Unique voice and communication style
- Proven workflows and success metrics
- Technical deliverables with code examples
Quick Start
Browse Available Agents
Agents are organized by domain. List available agents:
ls -la references/agents/Each .md file contains a complete agent personality with identity, mission, rules, and deliverables.
Major domains:
engineering/— Backend, Frontend, DevOps, Security, ML, etc.product/— Product Managers, UX Designers, Growth strategistssales/— Sales specialists, Account Executives, Deal closersmarketing/— Content, Paid Media, Community, Growthdesign/— UI/UX, Brand, Product Designstrategy/— Consultants, Analysts, Business strategistssupport/— Customer Success, Support specialiststesting/— QA Engineers, Test strategists- And more:
academic/,game-development/,integrations/,specialized/
Activate an Agent
- Find the agent in
references/agents/<domain>/(e.g.,engineering-frontend-developer.md) - Read the agent file to absorb personality, mission, and workflows
- Adopt the persona — use the agent's voice, approach, and technical style for your current task
- Follow their rules — each agent has critical guidelines and success metrics
Example: Frontend Developer
Read references/agents/engineering/engineering-frontend-developer.md, then work with:
- Their performance-first mindset
- Accessibility and inclusive design focus
- Technical deliverables (React components, PWAs, design systems)
- Communication style (detail-oriented, user-centric, precise)
Agent Structure
Each agent file contains:
---
name: Agent Name
description: Brief description
color: Visual indicator
emoji: 🎭
vibe: One-liner describing their energy
---
# Agent Personality
## 🧠 Identity & Memory
- Role, personality, experience
## 🎯 Core Mission
- Primary responsibilities
- Technical domains
- Key deliverables
## 🚨 Critical Rules
- Non-negotiable guidelines
- Quality standards
- Red lines
## 📋 Technical Deliverables
- Code examples
- Templates
- Proven patterns
## 💬 Communication Style
- Tone and voice
- How they interact
- Success metricsWorkflow: Task-Agent Matching
When facing a complex task:
- Identify the domain — What expertise does this need? (engineering, marketing, sales, etc.)
- Find the best agent — Browse that domain folder for the closest match
- Read their profile — Understand their perspective, priorities, and rules
- Adopt their approach — Use their workflows, communication style, and technical standards
- Execute — Apply their expertise and deliverables to your task
Example flow:
- Task: "Build a responsive dashboard"
- Domain: Engineering (Frontend)
- Agent:
engineering-frontend-developer.md - Adopt: Performance-first, accessibility-focused, Core Web Vitals optimization
- Execute: Build with their rules, code examples, and UX patterns
Tips for Effective Agent Use
🎯 Specificity Matters
- Don't just activate "Engineer" — pick a specific role like Frontend Developer, Backend Architect, or DevOps Automator
- The more specific the agent, the more tailored their expertise
📚 Read Actively
- Each agent file is self-contained and comprehensive
- Reading takes 2-3 minutes; the detail is worth it
- You'll pick up their mental models and priorities
🔄 Mix and Match
- You can combine insights from multiple agents (e.g., Frontend Developer + UX Designer for a UI project)
- Cross-domain personas often have complementary perspectives
✅ Follow Their Rules
- Each agent has non-negotiable guidelines ("Critical Rules")
- These aren't suggestions — they're their professional standards
- Following them ensures quality output
Available Agent Categories
| Domain | Focus | Example Agents |
|---|---|---|
| Engineering | 23 roles | Frontend Dev, Backend Architect, DevOps, Security, ML, etc. |
| Product | 8 roles | Product Manager, UX Designer, Growth, Analytics |
| Sales | 6 roles | Sales Rep, Account Exec, Deal Closer, Pipeline builder |
| Marketing | 8 roles | Content, Paid Media, Community, Growth, Brand |
| Design | 7 roles | UI/UX, Product Design, Brand, Design System |
| Strategy | 5 roles | Consultant, Analyst, Business Strategist |
| Support | 4 roles | Customer Success, Support, Community |
| Testing | 4 roles | QA Engineer, Test Strategist, Automation |
| Other | 12+ roles | Academic, Game Dev, Integrations, Specialized |
Reference Structure
All agent files are stored in references/agents/<domain>/ following the source repository structure. Each file is a standalone personality — load only the ones you need.
To see available agents, explore:
references/agents/engineering/— All engineering specialtiesreferences/agents/product/— Product and UX rolesreferences/agents/sales/— Sales professionals- (and so on for all domains)