Skip to main content

uag logo

uag (Local AI Agent)

uag is a local interactive agent that executes commands, manipulates files, and reads data files such as PDF, PPTX, and Excel. It provides three user interfaces: CLI, GUI, and Web. uag is built to keep you free from vendor-locked apps: use the interface that fits your workflow, switch providers, and stay in control of your environment. GitHub: https://github.com/awaku7/agentcli

Installation

Install from PyPI with pip:

pip install uag

If you use a virtual environment, activate it first and then run the command above.

On first launch, uag checks your environment and starts the setup wizard automatically when required provider variables are missing. For configuration details, see ENVIRONMENT.md.

Key Features

  • Practical toolset: File manipulation, web search, PDF/PPTX/Excel extraction, image generation, and image analysis.
  • Multi-provider support: OpenAI / Azure / Bedrock / OpenRouter / Ollama / Gemini / Vertex AI / Claude / Grok / NVIDIA.
  • Three interfaces:
    • CLI: uag / python -m uagent
    • GUI: uagg / python -m uagent.gui
    • Web: uagw / python -m uagent.web
    • A2A server: uaga / python -m uagent.a2a.server
  • MCP support: Connect to external MCP tool servers.
  • Session continuity: Keep context when switching models or providers.
  • Web Inspector: Save browser transitions, DOM snapshots, and screenshots with playwright_inspector.
  • Built-in docs: Read bundled docs with uag docs.

Usage

Start and exit

Run uag in your terminal to start. Type :exit to quit.

A2A server

Launch an Agent2Agent-compatible HTTP server:

uaga

See ENVIRONMENT.md for UAGENT_A2A_* settings such as auth, host, port, reload, public base URL, concurrency, and engine.

Handy tips

  • :tools: show loaded tools
  • :logs [n]: show recent session logs
  • :load <index>: load a previous session
  • :skills: select and load Agent Skills
  • :shrink [n]: summarize history and keep the last n messages

Configuration and details

Environment variables and setup

For API keys, language settings (UAGENT_LANG), history shrink settings, and more, see ENVIRONMENT.md.

  • Setup wizard: python -m uagent.setup_cli
  • Encrypted environment: use uag_envsec to encrypt .env as .env.sec
  • Update encrypted values: uag_envsec add --file .env.sec --key NAME --value VALUE

Responses API note

If you set UAGENT_RESPONSES=1, Responses API is used for supported providers: OpenAI / Azure / Bedrock / OpenRouter / Ollama. Gemini / Claude / Vertex AI use their native API paths and are not covered by Responses API. Image analysis via Responses is currently limited to OpenAI / Azure / Bedrock / OpenRouter. For other providers, uag falls back to the provider-specific or chat-completions path.

Developer docs and translations

Download files

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

Source Distribution

uag-0.5.4.tar.gz (3.2 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

uag-0.5.4-py3-none-any.whl (3.5 MB view details)

Uploaded Python 3

File details

Details for the file uag-0.5.4.tar.gz.

File metadata

  • Download URL: uag-0.5.4.tar.gz
  • Upload date:
  • Size: 3.2 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.5

File hashes

Hashes for uag-0.5.4.tar.gz
Algorithm Hash digest
SHA256 4b69fb369c1da965c1b0746c87d4480d9a1a5d41febfc58ef09b08e459202572
MD5 38db91662e4e9e3417ad4e1fb9eae733
BLAKE2b-256 4adc326e8594472ac780696db8dc31322975359c85860ddc5a8051ea6244b696

See more details on using hashes here.

File details

Details for the file uag-0.5.4-py3-none-any.whl.

File metadata

  • Download URL: uag-0.5.4-py3-none-any.whl
  • Upload date:
  • Size: 3.5 MB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.5

File hashes

Hashes for uag-0.5.4-py3-none-any.whl
Algorithm Hash digest
SHA256 5754c1e327f226774d7ed4c7fdc667ccc2dc7de94d1c014645fa08ad8cdf4abb
MD5 13216b476204a4a3192e467ba203ad58
BLAKE2b-256 79ebd0db6eecc035d9ce9302cfb228d4c56ff8db3079dccab75916809fb4cee0

See more details on using hashes here.

Release history Release notifications | RSS feed

Supported by

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