Skip to content

🚀 Welcome to Flock! Take Flight with Declarative AI Agents 🚀

Flock Banner Dynamic TOML Badge LinkedIn Bluesky

Tired of wrestling with paragraphs of prompt text just to get your AI agent to do one specific thing? 😫 Enter Flock, the agent framework that lets you ditch the prompt-palaver and focus on what you want your agents to achieve!

Flock takes a declarative approach. Think of it like ordering a pizza 🍕: you specify the toppings (your agent's input and output), not the 30 steps the chef needs to follow. Flock, powered by modern LLMs and a clever architecture, figures out the "how" for you.

Built with real-world deployment in mind, Flock integrates seamlessly with tools like Temporal for building robust, fault-tolerant, and scalable agent systems right out of the box.

✨ Why Join the Flock?

Traditional Agent Frameworks 😟 Flock Framework 🐤🐧🐓🦆
🤯 Prompt Nightmare ✅ Declarative Simplicity
Long, brittle, hard-to-tune prompts Clear input/output specs (with types!)
💥 Fragile & Unpredictable ⚡ Robust & Production-Ready
Single errors crash the system Fault-tolerant via Temporal integration
🧩 Monolithic & Rigid 🔧 Modular & Flexible
Hard to extend or modify logic Pluggable Evaluators, Modules, Routers
⛓️ Basic Chaining 🚀 Advanced Orchestration
Often just linear workflows Dynamic routing, parallel/batch runs

💡 Core Ideas

Flock's power comes from a few key concepts:

  1. Declarative Agents: Define agents by what they do (inputs/outputs), not how. Flock uses Evaluators to handle the underlying logic (be it LLM calls, rules, or custom code).
  2. Modular Components: Extend agent capabilities with pluggable Modules (for memory, metrics, output formatting, etc.) that hook into the agent's lifecycle without touching its core definition.
  3. Intelligent Workflows: Chain agents explicitly or use Routers (like LLM-based or Agent-based) for dynamic decision-making on the next step.
  4. Reliable Execution: Run locally for debugging or switch seamlessly to Temporal for production-grade fault tolerance, retries, and state management.
  5. Type Safety: Leverage Python type hints and Pydantic for clear contracts, validation, and easier integration.

🐥 Hello Flock! - A Quick Taste

Building your first agent is refreshingly simple:

import os
from flock.core import Flock, FlockFactory 


# --------------------------------
# Define the model
# --------------------------------
# Flock uses litellm to talk to LLMs
# Please consult the litellm documentation for valid IDs:
# https://docs.litellm.ai/docs/providers
MODEL = "openai/gpt-4o"


# --------------------------------
# Create the flock and context
# --------------------------------
# The flock is the place where all the agents are at home
flock = Flock(name="hello_flock", description="This is your first flock!", model=MODEL)

# --------------------------------
# Create an agent
# --------------------------------
# The Flock doesn't believe in prompts (see the docs for more info)
# The Flock just declares what agents get in and what agents produce
# my_presentation_agent takes in a topic and outputs a
# funny_title, fun_slide_headers and fun_slide_summaries
presentation_agent = FlockFactory.create_default_agent(
    name="my_presentation_agent",
    input="topic",
    output="fun_title, fun_slide_headers, fun_slide_summaries"
)
flock.add_agent(presentation_agent)


# --------------------------------
# Run the flock
# --------------------------------
# Tell the flock who the starting agent is and what input to give it
flock.run(
    start_agent=presentation_agent, 
    input={"topic": "A presentation about robot kittens"}
)

That's it! Flock handled turning your simple declaration into the necessary LLM interaction and gave you back a clean, typed Python object.

🗺️ Ready to Explore?

  • Dive deeper into the Flock ecosystem:

  • Getting Started: Your first flight – installation and basic usage.

  • Core Concepts: Understand the foundations – Agents, Declarative logic, Workflows.

  • Key Components: Learn about the building blocks – Evaluators, Modules, Tools.

  • Guides: Practical walkthroughs for common tasks like chaining agents.

  • Deployment: Take your Flock to production with Temporal.

🤝 Join the Flock Community!

  • Flock is actively developed and welcomes contributions.

  • Check out the code on GitHub

  • Report issues or suggest features.

  • Help us improve the documentation!

Let's build amazing, reliable AI agent systems together! 🚀