hammad-python

Harmoniously Accelerated Micro-Modules (for) Application Development

hammad-python is an ecosystem of wrappers with the sole purpose of defining opinionated resources that I myself use in my projects and day to day work when building with Python.

The ecosystem is split into 4 main sub-packages:

hammad-python-core : Contains "stdlib" like resources such as converters, types, caching, etc. All of these resources can easily be implemented from scratch, and are built soley for convenience.
hammad-python-data : Contains a Collection class / system that implements a simple vector / non vector database system.
hammad-python-genai : Contains a collection of patterns for using Generative AI models as well as implementing concepts like Agentic Reasoning.
hammad-python-http : HTTP / API related resources and utilities. Contains resources related to MCP, HTTP, OpenAPI & GraphQL Servers.

Installation

You can install any of the sub-packages, or the main package through pip or uv.

pip install hammad-python

Quickstart

Easily Use Language Models With Various Tools & Patterns

from ham.genai import Agent, LanguageModel
from ham.http import SearchClient

client = SearchClient()

def search_web(query: str) -> str:
    return client.search(query)

# create your agent
# agents have rich functionality, and use `litellm` & `instructor` under the hood
agent = Agent(
    instructions = "You are a helpful assistant who can search the web",
    tools = [search_web]
)

print(agent.run("What is the weather in Tokyo?"))

>>> AgentResponse:
city='Tokyo' temperature=25.0 description='Clear sky'

>>> Model: openai/gpt-4o-mini
>>> Steps: 2
>>> Output Type: Response
>>> Total Tool Calls: 1

Vectorized & Searchable Collections

from ham.data import Collection

collection = Collection(vector=True)

collection.add("Rocks")
collection.add("Cucumber")
collection.add("Dragon")
collection.add("Apple")

# query the collection
for result in collection.query("Fruit"):
    print(result.item)

>>> Apple
>>> Cucumber
>>> Dragon
>>> Rocks

HTTP & API Resources

from ham.http import function_server

# instantly launch a server by decorating a function
@function_server(path="/some-endpoint", auto_start=True)
def some_endpoint(name : str) -> str:
    return f"Hello, {name}!"

Now in a separate file:

import requests

print(
    requests.post(
        "http://0.0.0.0:8000/some-endpoint",
        json = {"name": "John"}
    ).json()
)

>>> {'result' : 'Hello, John!'}

Metadata

Release files for hammad-python 0.0.38

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for hammad-python 0.0.38
File	Size	Uploaded
hammad_python-0.0.38.tar.gz	434.1 kB	Jul 21, 2025	Details

Built distribution (wheel)

Table of built distributions (wheels) for hammad-python 0.0.38
File	Interpreter	ABI	Platform	Reset
hammad_python-0.0.38-py3-none-any.whl 5.2 kB Jul 21, 2025	Python 3	none	any	Details

Total release size: 439.3 kB

Release files / hammad_python-0.0.38.tar.gz

Download URL	hammad_python-0.0.38.tar.gz
Size	434.1 kB
Tags	Source
SHA-256 checksum How to use checksums	`d300a6c48daecdf60a99fb07e25576d4beb34d1a2ffeaceada5029d5a0dc22f0`
BLAKE2b-256 checksum How to use checksums	`1af6081360b06d72be9f922dc4d3e795e2b9ebf4e7bd615d2200653c5e114b2d`
Upload date	Jul 21, 2025
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.1.0 CPython/3.12.11`

Release files / hammad_python-0.0.38-py3-none-any.whl

Download URL	hammad_python-0.0.38-py3-none-any.whl
Size	5.2 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`e719ee198c0b17746c69468e1e54e11653d2e815e5928bc22adeac6ee997f63d`
BLAKE2b-256 checksum How to use checksums	`aed43019bacbccfa25264b57c5c48e61769d702d39bcc1b84ad470693d841aa0`
Upload date	Jul 21, 2025
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.1.0 CPython/3.12.11`