agentify-toolkit 0.4.3a1

Declarative AI toolkit

📦 Agentify - Declarative AI Agent Toolkit

Build and experiment with AI agents using simple declarative specs.

Agentify is a lightweight, declarative-first toolkit for prototyping AI agents. It lets you define agents as YAML specs and test them rapidly from the CLI or Python, without committing to a framework or model provider.

Use YAML for declarative specs, or Python for full control - both use the same agent model.

Note: Agentify is not a workflow orchestrator, runtime platform, or production framework. It’s for rapid Agent building, experimentation and prototyping.

Why Agentify ?

Most agent frameworks optimise for running agents - orchestration, routing, retries, supervision, cloud execution.

But before production comes a more important phase: figuring out what the agent should even do.

Agentify optimises for that phase.

Designed For:

✔ Rapid agent prototyping & ideation

✔ Teams exploring internal AI use cases

✔ Startups validating agent product ideas

✔ Teaching & education (low cognitive overhead)

✔ Provider portability (no early lock-in)

Quickstart

1. Install

pip install agentify-toolkit

2. Add an LLM Provider Key

# anthropic | openai | xai | google | bedrock
agentify providers add <provider>

Verify:

agentify providers list

Example Output:

 anthropic
  env: ANTHROPIC_API_KEY
  status: READY

3. Create an Agent (YAML)

name: claude
description: AI Engineer
version: 0.1.0
model:
  provider: anthropic
  id: claude-sonnet-4-5
  api_key_env: ANTHROPIC_API_KEY
role: |
  You are an AI Security Engineer.
  Provide concise, practical answers with examples.

4. Run the Agent

agentify run agent.yaml

You’ve just built your first AI agent with Agentify!

💡 Tip: Running multiple agents

If you have multiple agents, put them in a single folder and run agentify run <foldername>. Agentify will provide an interactive menu so you can choose which agent you want to experiment with.

💡 Tip: Overriding the model

If you want to experiment with a different model, simply add --provider=openai model=gpt-5-nano to your call. Ensure you have registered the appropriate provider API key.

agentify run agent.yaml --provider=openai --model=gpt-5-nano

Core Ideas

1. Declarative Agents (YAML-first)

Agents become artifacts:

• version controlled
• diffable
• shareable
• auditable

2. Provider Abstraction Without Lock-in

Most ecosystems ask:

“Am I building on OpenAI, Anthropic, Bedrock, XAI, or Google?”

Agentify flips it:

“The agent spec stays the same - only the provider changes.”

3. CLI-first Exploration

CLI interaction is fast, ergonomic, and repeatable:

agentify run agent.yaml

4. Agent = Single File

Agents collapse to a spec, not a codebase

Key Features

• Declarative agent definitions via YAML

• Multi-provider LLM support (OpenAI, Anthropic, XAI, Gemini, Bedrock)

• Interactive CLI and TUI for exploring agents

• Programmatic API for custom workflows

• Lightweight: Click + Rich + PyYAML

Documentation & Notebooks

Prefer a guided walkthrough?

• Developer Quickstart (Notebook) examples/notebooks/Agentify_Developer_Quickstart.ipynb

• YAML Deep Dive examples/notebooks/Agentify_YAML_Deep_Dive.ipynb

Programmatic Usage

from agentify import Agent

agent = Agent(
    name="Grok",
    description="X's Grok Agent",
    provider="x",
    model_id="grok-4",
    role="You are an AI Security Architect specialising in X AI Grok models"
)

response = agent.run("Which AI LLM is the best in 1 sentence?")
print(response)

CLI Reference

Action	Command
Run from YAML	agentify run agent.yaml
Run folder of agents	agentify run agents/
List agents interactively	agentify list agents
Add a provider API key	agentify providers add <p>
List provider credentials	agentify providers list

Supported Providers & Keys

Provider	Env Var
OpenAI	export OPENAI_API_KEY=...
Anthropic	export ANTHROPIC_API_KEY=...
Gemini	export GEMINI_API_KEY=...
XAI (Grok)	export XAI_API_KEY=...
Bedrock	export AWS_BEARER_TOKEN_BEDROCK

Windows:

$env:OPENAI_API_KEY="..."

Installation from PyPI:

pip install agentify-toolkit

From source:

git clone https://github.com/backplane-cloud/agentify-toolkit.git
cd agentify-toolkit
pip install .

License

Apache 2.0 - see LICENSE