Agentify-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.
Note: Agentify is not a workflow orchestrator or production framework. It’s simply for agent building, experimentation and prototyping.
Quickstart
For a more detailed step-by-step Quickstart.
1. Install the Agentify-Toolkit
pip install agentify-toolkit
2. Configure Provider API Keys
You can configure provider keys either via the CLI or manually through .env.
Option A - Using the CLI
agentify provider add <provider>
This updates (or creates) your .env file and stores the key for the selected provider.
To list all configured providers:
agentify provider list
To remove a provider key:
agentify provider remove <provider>
Option B - Using a .env File
cp .env.example .env
Populate .env with your provider keys:
OPENAI_API_KEY=<your-openai-key>
ANTHROPIC_API_KEY=<your-anthropic-key>
XAI_API_KEY=<your-xai-key>
GOOGLE_API_KEY=<your-google-key>
BEDROCK_API_KEY=<your-bedrock-key>
MISTRAL_API_KEY=<your-mistral-key>
DEEPSEEK_API_KEY=<your-deepseek-key>
OLLAMA_API_KEY=<your-deepseek-key>
Any configured provider will be automatically detected at runtime.
For instructions on how to obtain an Model API key:
| Provider | Model | Link |
|---|---|---|
| OpenAI | GPT-4 | How to obtain an OpenAI API Key |
| Gemini | How to obtain an Google API Key | |
| Anthropic | Claude | How to obtain an Anthropic API Key |
| XAI | Grok | How to obtain an XAI API Key |
| Mistral AI | Mistral | How to obtain an Mistra AI API Key |
| Deepseek | deepseek-chat | How to obtain an Deepseek API Key |
| Ollama | evstral-small-2:24b | How to obtain an Ollama Cloud API Key |
Verify:
agentify provider list
Example Output:
Configured Providers:
✓ openai (sk-s****)
✓ anthropic (sk-a****)
✓ deepseek (sk-5****)
✓ mistral (XOsY****)
✓ xai (xai-****)
✓ google (AIza****)
✓ bedrock (ABSK****)
✓ ollama (4163****)
3. Create an Agent
You can generate an agent spec via the CLI:
agentify agent new
Or define one manually by creating 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
Run an agent directly from its YAML spec:
agentify run agent.yaml
You’ve just built and executed your first AI agent with Agentify.
Running Multiple Agents
You’ve just built and executed your first AI agent with Agentify.
agentify run examples/agents
Agentify will present an interactive selector so you can choose which agent to execute.
Overriding the Model at Runtime
Models and providers can be swapped without editing the YAML. For example:
agentify run agent.yaml --provider=openai --model=gpt-5-nano
Using overrides is useful for experimentation or benchmarking. Ensure the required API key is configured.
Programmatic Usage
from agentify import Agent
# Create 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"
)
# Sent Prompt
response = agent.run("Which AI LLM is the best in 1 sentence?")
print(response)
Quick CLI Reference
| Action | Command |
|---|---|
| Run from YAML | agentify run agent.yaml |
| Run folder of agents | agentify run examples/agents |
| List agents interactively | agentify agent list [<folder_name>] |
| Add a provider API key | agentify provider add <p> |
| List provider credentials | agentify provider 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
Install 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
1 maintainer
Release files for agentify-toolkit 0.22.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| agentify_toolkit-0.22.0.tar.gz | 75.4 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| agentify_toolkit-0.22.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 168.9 kB
Release files / agentify_toolkit-0.22.0.tar.gz
| Download URL | agentify_toolkit-0.22.0.tar.gz |
|---|---|
| Size | 75.4 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
1104940138d49a4ec266be6f7ed3a54c5a1c0b84266d2d57eb97ba4bd8df68d8
|
|
BLAKE2b-256 checksum How to use checksums |
9d87c016cbe4601d601fd69d578e58138da30a7a13dce4e8713b153cf5b51878
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.14.0
|
Release files / agentify_toolkit-0.22.0-py3-none-any.whl
| Download URL | agentify_toolkit-0.22.0-py3-none-any.whl |
|---|---|
| Size | 93.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
c42e4e7640bc188d39d83069e54564332d3c0d1c519ca8dc0d93b4c7d244ac59
|
|
BLAKE2b-256 checksum How to use checksums |
ddf163ab96e7519ce87863007a55ff7950d22c604552b269fbada39fc3451119
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.14.0
|
