IO Intelligence Agent Framework

IMPORTANT: Beta Notice: This project is in rapid development and may not be stable for production use.

This repository provides a flexible system for building and orchestrating agents and workflows. It offers two modes:

• Client Mode: Where tasks call out to a remote API client (e.g., your client.py functions).
• Local Mode: Where tasks run directly in the local environment, utilizing run_agents(...) and local logic.

It also supports loading YAML or JSON workflows to define multi-step tasks.

Table of Contents

1. Overview
2. Installation
3. Concepts
   • Agents
   • Tasks
   • Client Mode vs Local Mode
   • Workflows (YAML/JSON)
4. Usage
   • Creating Agents
   • Creating an Agent with custom Persona
   • Building a Workflow
   • Running a Local Workflow
   • Running a Remote Workflow (Client Mode)
   • Uploading YAML/JSON Workflows
5. Examples
   • Simple Summarize Task
   • Chainable Workflows
   • Custom Workflow
   • Loading From a YAML File
6. API Endpoints
7. License

Overview

The framework has distilled Agents into 3 distinct pieces:

• Agents
• Tasks
• Workflows

The Agent can be configured with:

• Model Provider (e.g., OpenAI, Llama, etc.)
• Tools (e.g., specialized functions)

Users can define tasks (like sentiment, translate_text, etc.) in a local or client mode. They can also upload workflows (in YAML or JSON) to orchestrate multiple steps in sequence.

Installation

1. Install the latest release:

pip install --upgrade iointel

2. For UI features (Gradio interface), install with UI dependencies:

pip install --upgrade "iointel[ui]"

3. Set Required Environment Variable:

   • OPENAI_API_KEY or IO_API_KEY for the default OpenAI-based ChatOpenAI.

4. Optional Environment Variables:

   • AGENT_LOGGING_LEVEL (optional) to configure logging verbosity: DEBUG, INFO, etc.
   • OPENAI_API_BASE_URL or IO_API_BASE_URL to point to OpenAI-compatible API implementation, like https://api.intelligence.io.solutions/api/v1
   • OPENAI_API_MODEL or IO_API_MODEL to pick specific LLM model as "agent brain", like openai/gpt-oss-120b

Concepts

Agents

• They can have a custom model (e.g., OpenAIModel, a Llama-based model, etc.).
• Agents can have tools attached, which are specialized functions accessible during execution.
• Agents can have a custom Persona Profile configured.

Tasks

• A task is a single step in a workflow, e.g., schedule_reminder, sentiment, translate_text, etc.
• Tasks are managed by the Workflow class in workflow.py.
• Tasks can be chained for multi-step logic into a workflow (e.g., await Workflow(objective="...").translate_text().sentiment().run_tasks()).

Client Mode vs Local Mode

• Local Mode: The system calls run_agents(...) directly in your local environment.
• Client Mode: The system calls out to remote endpoints in a separate API.
  • In client_mode=True, each task (e.g. sentiment) triggers a client function (sentiment_analysis(...)) instead of local logic.

This allows you to switch between running tasks locally or delegating them to a server.

Workflows (YAML/JSON)

Note: this part is under active development and might not always function!

• You can define multi-step workflows in YAML or JSON.
• The endpoint /run-file accepts a file (via multipart form data).
  • First tries parsing the payload as JSON.
  • If that fails, it tries parsing the payload as YAML.
• The file is validated against a WorkflowDefinition Pydantic model.
• Each step has a type (e.g., "sentiment", "custom") and optional parameters (like agents, target_language, etc.).

Usage

Creating Agents

from iointel import Agent

my_agent = Agent(
    name="MyAgent",
    instructions="You are a helpful agent.",
    # one can also pass custom model using pydantic_ai.models.openai.OpenAIModel
    # or pass args to OpenAIModel() as kwargs to Agent()
)

Creating an Agent with a Persona

from iointel import PersonaConfig, Agent


my_persona = PersonaConfig(
    name="Elandria the Arcane Scholar",
    age=164,
    role="an ancient elven mage",
    style="formal and slightly archaic",
    domain_knowledge=["arcane magic", "elven history", "ancient runes"],
    quirks="often references centuries-old events casually",
    bio="Once studied at the Grand Academy of Runic Arts",
    lore="Elves in this world can live up to 300 years",
    personality="calm, wise, but sometimes condescending",
    conversation_style="uses 'thee' and 'thou' occasionally",
    description="Tall, silver-haired, wearing intricate robes with arcane symbols",
    emotional_stability=0.85,
    friendliness=0.45,
    creativity=0.68,
    curiosity=0.95,
    formality=0.1,
    empathy=0.57,
    humor=0.99,
)

agent = Agent(
    name="ArcaneScholarAgent",
    instructions="You are an assistant specialized in arcane knowledge.",
    persona=my_persona
)

print(agent.instructions)

Building a Workflow

In Python code, you can create tasks by instantiating the Tasks class and chaining methods:

from iointel import Workflow

tasks = Workflow(objective="This is the text to analyze", client_mode=False)
(
  tasks
    .sentiment(agents=[my_agent])
    .translate_text(target_language="french")   # a second step
)

results = await tasks.run_tasks()
print(results)

Because client_mode=False, everything runs locally.

Running a Local Workflow

tasks = Workflow(objective="Breaking news: local sports team wins!", client_mode=False)
await tasks.summarize_text(max_words=50).run_tasks()

Running a Remote Workflow (Client Mode)

tasks = Workflow(objective="Breaking news: local sports team wins!", client_mode=True)
await tasks.summarize_text(max_words=50).run_tasks()

Now, summarize_text calls the client function (e.g., summarize_task(...)) instead of local logic.

Uploading YAML/JSON Workflows

Note: this part is under active development and might not always function!

1.	Create a YAML or JSON file specifying workflow:

name: "My YAML Workflow"
text: "Large text to analyze"
workflow:
  - type: "sentiment"
  - type: "summarize_text"
    max_words: 20
  - type: "moderation"
    threshold: 0.7
  - type: "custom"
    name: "special-step"
    objective: "Analyze the text"
    instructions: "Use advanced analysis"
    context:
      extra_info: "some metadata"

2.	Upload via the /run-file endpoint (multipart file upload).

The server reads it as JSON or YAML and runs the tasks sequentially in local mode.

Examples

Simple Summarize Task

tasks = Workflow("Breaking news: new Python release!", client_mode=False)
await tasks.summarize_text(max_words=30).run_tasks()

Returns a summarized result.

Chainable Workflows

tasks = Workflow("Tech giant acquires startup for $2B", client_mode=False)
(tasks
   .translate_text(target_language="spanish")
   .sentiment()
)
await results = tasks.run_tasks()

1.	Translate to Spanish,
2.	Sentiment analysis.

Custom Workflow

tasks = Workflow("Analyze this special text", client_mode=False)
tasks.custom(
    name="my-unique-step",
    objective="Perform advanced analysis",
    instructions="Focus on entity extraction and sentiment",
    agents=[my_agent],
    **{"extra_context": "some_val"}
)
await results = tasks.run_tasks()

A "custom" task can reference a custom function in the CUSTOM_WORKFLOW_REGISTRY or fall back to a default behavior.

Loading From a YAML File

Note: this part is under active development and might not always function!

curl -X POST "https://api.intelligence.io.solutions/api/v1/workflows/run-file" \
     -F "yaml_file=@path/to/workflow.yaml"

API Endpoints

Please refer to (IO.net documentation)[https://docs.io.net/docs/exploring-ai-agents] to see particular endpoints and their documentation.

License

See the LICENSE file for license rights and limitations (Apache 2.0).

IOIntel: Agentic Tools with Beautiful UI

Features (MISSING ANCHOR)

• Agentic tool use: Agents can call Python tools, return results, and chain reasoning.
• Rich tool call visualization: Tool calls and results are rendered as beautiful, gold-accented "pills" in both CLI (with rich) and Gradio UI.
• Dynamic UI: Agents can generate forms (textboxes, sliders, etc.) on the fly in the Gradio app.
• Live CSS theming: Agents can change the UI theme at runtime.
• Jupyter compatible: The Gradio UI can be launched in a notebook cell.

Quickstart: CLI Usage (MISSING ANCHOR)

from iointel import Agent, register_tool

[@register_tool](https://github.com/register_tool)
def add(a: float, b: float) -> float:
    return a + b

agent = Agent(
    name="Solar",
    instructions="You are a helpful assistant.",
    model="gpt-4o",
    api_key="sk-...",
    tools=[add],
    show_tool_calls=True,  # Pretty rich tool call output!
)

import asyncio
async def main():
    result = await agent.run("What is 2 + 2?", pretty=True)
    # Tool calls/results are shown in rich formatting!

asyncio.run(main())

Multimodal Support (MISSING ANCHOR)

iointel supports multimodal inputs through various content types:

from iointel import Agent, ImageUrl, BinaryContent, DocumentUrl, AudioUrl, VideoUrl

agent = Agent(
    name="VisionAgent",
    instructions="You are a helpful vision assistant.",
    model="openai/gpt-oss-120b",
    api_key="io-...",
)

# Images
result = await agent.run([
    "What's in this image?",
    ImageUrl(url="https://example.com/image.png")
])

# Local images with binary content
with open("local_image.png", "rb") as f:
    image_data = f.read()

result = await agent.run([
    "Describe this image",
    BinaryContent(data=image_data, media_type="image/png")
])

# Documents
result = await agent.run([
    "Summarize this document",
    DocumentUrl(url="https://example.com/document.pdf")
])

# Audio/Video (model dependent)
result = await agent.run([
    "Transcribe this audio",
    AudioUrl(url="https://example.com/audio.mp3")
])

Supported Media Types: The specific media types supported depend on your LLM model provider:

• Images: PNG, JPEG, GIF, WebP
• Documents: PDF, TXT
• Audio/Video: MP3, MP4, WAV (varies by provider)

Check your model provider's documentation for specific format support and limitations.

Quickstart: Gradio UI (MISSING ANCHOR)

Note: To use the Gradio UI, install with UI dependencies: pip install "iointel[ui]"

from iointel import Agent, register_tool

[@register_tool](https://github.com/register_tool)
def get_weather(city: str) -> dict:
    return {"temp": 72, "condition": "Sunny"}

agent = Agent(
    name="GradioSolar",
    instructions="You are a helpful assistant.",
    model="openai/gpt-oss-120b",
    api_key="io-...",
    tools=[get_weather],
    show_tool_calls=True,
)

# Launch the beautiful Gradio Chat UI (works in Jupyter too!)
await agent.launch_chat_ui(interface_title="Iointel Gradio Solar")

# Or, for more control across different agents:
# from iointel.src.ui.io_gradio_ui import IOGradioUI
# ui = IOGradioUI(agent, interface_title="Iointel GradioSolar")
# await ui.launch(share=True)

• Tool calls are rendered as beautiful, gold-trimmed panels in the chat.
• Dynamic UI: If your agent/tool returns a UI spec, it will be rendered live.
• Works in Jupyter: Just run the above in a notebook cell!