Skip to main content

Integrate your OS with an LLM.

Project description

AIOS: give a LLM full access to your terminal for the sake of science.

This is basically https://github.com/VictorTaelin/AI-scripts/blob/main/chatsh.mjs implemented in Python, with additional features like image analysis, configuration files.

Example usage:

https://github.com/piEsposito/ai-os/assets/47679710/0c8bf030-963b-4a46-aae5-0b3925ed7c36

Installation

  1. Clone this repository.
  2. Run poetry install in the project directory.
  3. Alternatively, install via pip: pip install pi-ai-os

Usage

aios [options] [initial message]

Options:

  • -m, --model: Specify the AI model to use (default: claude-3-5-sonnet-20240620)
  • -c, --config: Path to the configuration file
  • -a, --assistant: Use an assistant model for context gathering, hardcoded to Claude Haiku for speed and cost
  • -v, --verbose: Increase output verbosity

Example:

export ANTHROPIC_API_KEY=your_api_key_here
aios -m claude-3-5-sonnet-20240620 "List all files in the current directory"

Features

Image Analysis

To include an image in your query, use the /image command followed by the path to the image:

/image path/to/your/image.jpg Describe this image

Configuration File

AIOS can use a configuration file to provide context about your project. By default, it looks for a file named pi_ai_os.yaml in the current directory. You can specify a different config file using the -c option.

Example configuration file:

allowed_extensions:
  - .txt
  - .md
  - .py
ignored_dirs:
  - .git
  - node_modules

Context-Aware Assistance

Use the -a or --assistant flag to enable context-aware assistance. This feature uses an assistant model to analyze the project structure and select relevant files for providing context to the main AI model.

Clearing Context

Type /clear or /c during your interaction with AIOS to clear the conversation context.

Verbose Mode

Use the -v or --verbose flag to increase output verbosity, which can be helpful for debugging or understanding the AI's decision-making process.

Project details


Download files

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

Source Distribution

pi_ai_os-0.0.4.tar.gz (9.5 kB view details)

Uploaded Source

Built Distribution

pi_ai_os-0.0.4-py3-none-any.whl (10.6 kB view details)

Uploaded Python 3

File details

Details for the file pi_ai_os-0.0.4.tar.gz.

File metadata

  • Download URL: pi_ai_os-0.0.4.tar.gz
  • Upload date:
  • Size: 9.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.0 CPython/3.12.4

File hashes

Hashes for pi_ai_os-0.0.4.tar.gz
Algorithm Hash digest
SHA256 e19a21952ebe4951b272583447827245a18999baa557d6ede192fe1630d23cdb
MD5 e324c8476f783e7bd3a52940de285716
BLAKE2b-256 116d3e6b74358c29c34c47637d909b5c32808cd99c9c00f4feaa5688571d0295

See more details on using hashes here.

File details

Details for the file pi_ai_os-0.0.4-py3-none-any.whl.

File metadata

  • Download URL: pi_ai_os-0.0.4-py3-none-any.whl
  • Upload date:
  • Size: 10.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/5.1.0 CPython/3.12.4

File hashes

Hashes for pi_ai_os-0.0.4-py3-none-any.whl
Algorithm Hash digest
SHA256 e9477e09eebafc2250b33d6b17fe3062a121642f33ea6c9145530c18c1da22cb
MD5 cb1202dc487a9237c3ca3ba8b136d286
BLAKE2b-256 cc452bdd08721421e7cd23316f823175fac918bdefdcb1269d94d4e19865f421

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