Agency-Agents¶

What it is¶

Agency-Agents is a comprehensive suite of 110+ specialized AI agent personas designed to transform generic coding assistants into a "complete AI agency." It provides markdown-based system prompts for roles ranging from Frontend Architects to Security Engineers and Reality Checkers.

What problem it solves¶

It reduces AI hallucinations and improves code quality by providing "off-the-shelf" expert personas with predefined identities, missions, and success metrics. It moves away from generic "write code" prompts to specialized, opinionated domain expertise.

Where it fits in the stack¶

Category: Agent / Personas / Framework. It acts as a configuration layer for IDE-based agents like Claude Code, Cursor, and Aider.

Typical use cases¶

Multi-agent IDE Workflows: Invoking a "Backend Architect" for design and a "Security Engineer" for review.
Reality Checking: Using the "Reality Checker" persona to find flaws in proposed solutions.
Specialized Automation: Deploying "Reddit community ninjas" or "Whimsy injectors" for niche business tasks.

Getting started¶

Installation¶

# Clone the repository to your local machine
git clone https://github.com/msitarzewski/agency-agents.git ~/.agency-agents

Integration with Claude Code¶

To use a persona with Claude Code, you can reference the markdown file in your system prompt or initial query:

claude "Use the instructions in ~/.agency-agents/agents/backend-architect.md to design a FastAPI service."

CLI examples¶

# List all available agent personas
ls ~/.agency-agents/agents/

# Use the 'Security Engineer' persona with Aider
aider --model claude-3-5-sonnet-20241022 --message-file ~/.agency-agents/agents/security-engineer.md

# Search for a specific specialist (e.g., Frontend)
ls ~/.agency-agents/agents/ | grep "frontend"

API examples¶

While primarily markdown-based, you can use these personas in your own Python agents:

import os

def load_persona(agent_name):
    path = os.path.expanduser(f"~/.agency-agents/agents/{agent_name}.md")
    with open(path, "r") as f:
        return f.read()

# Load the 'Reality Checker' persona
reality_checker_prompt = load_persona("reality-checker")

# Use it with your favorite LLM client
# response = client.chat(messages=[{"role": "system", "content": reality_checker_prompt}, ...])

Strengths¶

High Specialization: 110+ personas covering almost every development and business role.
Model Agnostic: Works with any LLM (Claude, GPT, Gemini) through any interface.
Improved Grounding: Drastically reduces hallucinations by narrowing the agent's focus.

Limitations¶

Manual Integration: Requires cloning and manual referencing in most tools.
Persona Maintenance: Prompts may need adjustment as models evolve.

When to use it¶

When you need more than just a general-purpose assistant and want a "team" of experts.
For business workflows that map well to traditional agency roles.

When not to use it¶

For simple tasks where a single general assistant (e.g., Claude or ChatGPT) is sufficient.

Sources / References¶

Contribution Metadata¶

Last reviewed: 2026-05-19
Confidence: high