Skip to main content

An AI-powered console assistant with a versatile API for seamless integration into applications.

Project description

Description or Alt text

License: MIT PyPI Docker PyPI

Build highly capable Single and Multi-Agent Systems.

BondAI Homepage

Checkout the BondAI Homepage (https://bondai.dev) for in depth documentation, examples and API specification.

Getting Started

There are 3 ways to use BondAI:

  1. 🛠️ Command Line Interface (CLI) - This is the easiest way to get up and running fast. Run BondAI on your command line with a pre-configured set of tools.

  2. 🐋 Docker - Running BondAI in a Docker container is recommended if you plan on using tools that run code or directly access your shell.

  3. 🏗️ Start Coding with BondAI - Integrate BondAI into your own codebase and start building your own agents.

🚀 Installation

Installing BondAI is easy:

pip install bondai

🛠️ Command Line Interface (CLI)

Once you've installed BondAI using pip install bondai the CLI will be available as an executable accessible simply by running bondai in your shell. Learn more about all of the tools and options available through the CLI.

Before running bondai you will need to set the OPENAI_API_KEY environment variable.

export OPENAI_API_KEY=sk-XXXXXXXXXX

Once the environment variable has been set you can run bondai to start the CLI.

Loading BondAI...
******************ENTERING CHAT******************
You are entering a chat with BondAI...
You can exit any time by typing 'exit'.

Hello! I'm BondAI, your friendly and helpful assistant. I'm here to assist you with any tasks or questions you might have. How can I assist you today?

I want you to write a story about unicorns and save it to a file named unicorns.md.
Using tool file_write: Writing a story about unicorns and saving it to a file named unicorns.md
Using tool final_answer...

A story about unicorns has been successfully written and saved to a file named unicorns.md. The story is set in an enchanted forest and describes the magical and majestic nature of unicorns, their daily routines, and their harmonious relationship with other creatures in the forest.

🐋 Docker

BondAI Docker images are available on DockerHub here. If you intend to use tools that run arbitrary code (PythonREPLTool) or access your shell (ShellTool) it is highly recommended that you run BondAI in a Docker container as these tools can damage your machine.

Before running the BondAI Docker container it is recommended that you create a directory named 'agent-volume' and mount it as a volume on the container. This will be used as the Agent's working directory and allows you to easily share files with the Agent.

mkdir agent-volume
docker pull krohling/bondai:latest
docker run -it --rm \
           -v ./agent-volume:/agent-volume \
           -w /agent-volume \
           OPENAI_API_KEY=sk-XXXXXXXXXX \
           bondai:latest bondai

🔥 Start Coding with BondAI

BondAI has a straightforward API for creating powerful AI Agents. Check out our examples for ideas on how to get started. Remember to set your OPENAI_API_KEY environment variable before running your BondAI Agent.

from bondai.agents import Agent
from bondai.tools.search import DuckDuckGoSearchTool
from bondai.tools.website import WebsiteQueryTool
from bondai.tools.file import FileWriteTool

task = """I want you to research the usage of Metformin as a drug to treat aging and aging related illness. 
You should only use reputable information sources, ideally peer reviewed scientific studies. 
I want you to summarize your findings in a document named metformin.md and includes links to reference and resources you used to find the information. 
Additionally, the last section of your document you should provide a recommendation for a 43 year old male, in good health and who regularly exercises as to whether he would benefit from taking Metformin. 
You should explain your recommendation and justify it with sources. 
Finally, you should highlight potential risks and tradeoffs from taking the medication."""

Agent(tools=[
  DuckDuckGoSearchTool(),
  WebsiteQueryTool(),
  FileWriteTool()
]).run(task)

BondAI Integrations

BondAI comes out of the box with a powerful set of integrations.

openai logo OpenAI BondAI supports any combination of OpenAI models and services including GPT-4, GPT-3.5, Dalle-E 3, and Embeddings.
azure logo Microsoft Azure BondAI fully supports connectivity to GPT-N, Dalle-E and Embedding APIs through Microsoft's Azure OpenAI services.
google logo Google Search Allows BondAI to search the internet. Requires a Google Search API Key and CSE ID
duckduckgo logo DuckDuckGo Allows BondAI to search the internet. No API keys required.
alpaca markets logo Alpaca Markets Allows BondAI to buy and sell stocks and crypto. Requires an Alpaca Markets account.
postgres logo PostgreSQL BondAI can automatically extract the schema from a Postgres DB and process natural language queries.
bland.ai logo Bland AI Allows BondAI to make phone calls and process/retrieve call transcripts. Requires a Bland.ai account.
gmail logo Gmail Allows BondAI to search and read emails.

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

bondai-0.3.1.tar.gz (26.7 MB view details)

Uploaded Source

Built Distribution

bondai-0.3.1-py3-none-any.whl (103.4 kB view details)

Uploaded Python 3

File details

Details for the file bondai-0.3.1.tar.gz.

File metadata

  • Download URL: bondai-0.3.1.tar.gz
  • Upload date:
  • Size: 26.7 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.12.1

File hashes

Hashes for bondai-0.3.1.tar.gz
Algorithm Hash digest
SHA256 7bf09003f64a0cde505b6567bc3a79b1c1ece8e35689360c425b7677ec5a698e
MD5 26d02f095b197360e78eb6aad6506005
BLAKE2b-256 44b9be3fb0308dd61f91052fcb1238adb3f524876c9d36a4f1aed8570f8b6c01

See more details on using hashes here.

File details

Details for the file bondai-0.3.1-py3-none-any.whl.

File metadata

  • Download URL: bondai-0.3.1-py3-none-any.whl
  • Upload date:
  • Size: 103.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.12.1

File hashes

Hashes for bondai-0.3.1-py3-none-any.whl
Algorithm Hash digest
SHA256 02334e2de5eecb7de595f66df6b5622eb6badb881162fbeddd488cd8cf27bd88
MD5 0cf317626950f55e6660acea3829bd91
BLAKE2b-256 7a1ce5c2011e96a8563c94496f124443c8ddffb3f7006cb11222d68980c127aa

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page