# Agentia: Ergonomic LLM Agents
Ergonomic LLM Agents, with chat messages fully compatible with [Vercel AI SDK](https://ai-sdk.dev/).
# Getting Started
Run agents with tools and MCP.
```python
from agentia import Agent, MCPServer, MCPContext
from typing import Annotated
# Define a tool as a python function
def get_weather(location: Annotated[str, "The city name"]):
"""Get the current weather in a given location"""
return { "temperature": 72 }
# Declare a MCP server:
calc = MCPServer(name="calculator", command="uvx", args=["mcp-server-calculator"])
# Create an agent
agent = Agent(model="openai/gpt-5-mini", tools=[get_weather, calc])
# Run the agent with the tool
async with MCPContext(): # This line can be omitted if not using MCP
response = await agent.run("What is the weather like in boston?")
print(response.text)
# Output: The current temperature in Boston is 72°F.
```
# The Magic Decorator
Create agent-powered magic functions.
Support both plain types and pydantic models as input and output.
```python
from agentia import magic
from pydantic import BaseModel
class Forcast(BaseModel):
location: str
temperature_celsius: int
@magic
async def get_weather(weather_forcast: str) -> Forcast:
"""Create weather forcase object based on the input string"""
...
forcast = await get_weather("The current temperature in Boston is 72°F")
print(forcast.location) # Output: Boston
print(forcast.temperature_celsius) # Output: 22
```
# Supported Parameter and Result Types
* Any types that can be passed to `pydantic.TypeAdaptor`:
* Builtin types: `int`, `float`, `str`, `bool`, `tuple[_]`, `list[_]`, `dict[_, _]`
* Enums: `Literal['A', 'B', ...]`, `StrEnum`, `IntEnum`, and `Enum`
* dataclasses
* `pydantic.BaseModel` subclasses
# Run agent as a REPL app
1. Create a config file at `./robo.toml`
```toml
[agent]
name = "Robo" # This is the only required field
icon = "🤖"
instructions = "You are a helpful assistant"
model = "openai/o3-mini"
plugins = ["clock"]
[mcp]
calc={ command = "uvx", args = ["mcp-server-calculator"] }
```
2. Load the agent
```python
agent = Agent.from_config("./robo.toml")
```
Raw data
{
"_id": null,
"home_page": null,
"name": "agentia",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.12",
"maintainer_email": null,
"keywords": "agent, ai-sdk, chatgpt, gpt, llm, openai, openrouter, tools",
"author": null,
"author_email": "Wenyu Zhao <wenyuzhaox@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/df/5b/0be573dffb94585c1359b413791ff21f911af102a5326d16fef753ef857d/agentia-0.1.10.tar.gz",
"platform": null,
"description": "# Agentia: Ergonomic LLM Agents\n\nErgonomic LLM Agents, with chat messages fully compatible with [Vercel AI SDK](https://ai-sdk.dev/).\n\n# Getting Started\n\nRun agents with tools and MCP.\n\n```python\nfrom agentia import Agent, MCPServer, MCPContext\nfrom typing import Annotated\n\n# Define a tool as a python function\ndef get_weather(location: Annotated[str, \"The city name\"]):\n \"\"\"Get the current weather in a given location\"\"\"\n return { \"temperature\": 72 }\n\n# Declare a MCP server:\ncalc = MCPServer(name=\"calculator\", command=\"uvx\", args=[\"mcp-server-calculator\"])\n\n# Create an agent\nagent = Agent(model=\"openai/gpt-5-mini\", tools=[get_weather, calc])\n\n# Run the agent with the tool\nasync with MCPContext(): # This line can be omitted if not using MCP\n response = await agent.run(\"What is the weather like in boston?\")\n\nprint(response.text)\n\n# Output: The current temperature in Boston is 72\u00b0F.\n```\n\n# The Magic Decorator\n\nCreate agent-powered magic functions.\n\nSupport both plain types and pydantic models as input and output.\n\n```python\nfrom agentia import magic\nfrom pydantic import BaseModel\n\nclass Forcast(BaseModel):\n location: str\n temperature_celsius: int\n\n@magic\nasync def get_weather(weather_forcast: str) -> Forcast:\n \"\"\"Create weather forcase object based on the input string\"\"\"\n ...\n\nforcast = await get_weather(\"The current temperature in Boston is 72\u00b0F\")\n\nprint(forcast.location) # Output: Boston\nprint(forcast.temperature_celsius) # Output: 22\n```\n\n# Supported Parameter and Result Types\n\n* Any types that can be passed to `pydantic.TypeAdaptor`:\n * Builtin types: `int`, `float`, `str`, `bool`, `tuple[_]`, `list[_]`, `dict[_, _]`\n * Enums: `Literal['A', 'B', ...]`, `StrEnum`, `IntEnum`, and `Enum`\n * dataclasses\n* `pydantic.BaseModel` subclasses\n\n# Run agent as a REPL app\n\n1. Create a config file at `./robo.toml`\n\n```toml\n[agent]\nname = \"Robo\" # This is the only required field\nicon = \"\ud83e\udd16\"\ninstructions = \"You are a helpful assistant\"\nmodel = \"openai/o3-mini\"\nplugins = [\"clock\"]\n\n[mcp]\ncalc={ command = \"uvx\", args = [\"mcp-server-calculator\"] }\n```\n\n2. Load the agent\n\n```python\nagent = Agent.from_config(\"./robo.toml\")\n```\n",
"bugtrack_url": null,
"license": null,
"summary": "Ergonomic LLM Agents",
"version": "0.1.10",
"project_urls": {
"Repository": "https://github.com/wenyuzhao/agentia"
},
"split_keywords": [
"agent",
" ai-sdk",
" chatgpt",
" gpt",
" llm",
" openai",
" openrouter",
" tools"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "3186259cbee1fc60a99756dffe4e638e508c4de2a9f3a7ecb8def6efc53b2e75",
"md5": "d8a97b7eab5d90603e4dbb6004ba8a79",
"sha256": "a806ea35f48dc3bd6e305f3b1ccc0c8ecfab4668c5d25f190213a8f6f98aa8b9"
},
"downloads": -1,
"filename": "agentia-0.1.10-py3-none-any.whl",
"has_sig": false,
"md5_digest": "d8a97b7eab5d90603e4dbb6004ba8a79",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.12",
"size": 41661,
"upload_time": "2025-10-26T07:07:49",
"upload_time_iso_8601": "2025-10-26T07:07:49.266338Z",
"url": "https://files.pythonhosted.org/packages/31/86/259cbee1fc60a99756dffe4e638e508c4de2a9f3a7ecb8def6efc53b2e75/agentia-0.1.10-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "df5b0be573dffb94585c1359b413791ff21f911af102a5326d16fef753ef857d",
"md5": "f8b81b15de140cc75bab515b1d8efb13",
"sha256": "892c49687507260ec71669128e4f0c6f6d0ca9c36af36147477c4d4e4db41dfa"
},
"downloads": -1,
"filename": "agentia-0.1.10.tar.gz",
"has_sig": false,
"md5_digest": "f8b81b15de140cc75bab515b1d8efb13",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.12",
"size": 31843,
"upload_time": "2025-10-26T07:07:50",
"upload_time_iso_8601": "2025-10-26T07:07:50.896540Z",
"url": "https://files.pythonhosted.org/packages/df/5b/0be573dffb94585c1359b413791ff21f911af102a5326d16fef753ef857d/agentia-0.1.10.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-10-26 07:07:50",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "wenyuzhao",
"github_project": "agentia",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "agentia"
}
JSON
