OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Frameworks

Stirrup

Stirrup is a lightweight Python framework and starting-point template for building customizable AI agents. It supplies an agent loop, session and tool abstractions, and built-in capabilities such as code execution, web search, document handling and context management.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Stirrup — OpenViST project profile in Agent Frameworks.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Stirrup when building a Python agent in code and you want a customizable starting point with tool and session abstractions rather than a finished user-facing assistant.

Who it is for

  • Python developers building a custom AI agent.
  • Teams that want an agent framework with built-in tools and the option to add their own.
  • Developers who want to choose among OpenAI-compatible clients, LiteLLM or a custom client.

Documented uses

  • Create an agent whose model selects how to approach a task.
  • Give an agent tools for code execution, web search, page fetching or MCP servers.
  • Import documents into agent context and produce file outputs.
  • Extend an agent with modular skills or custom tools.
  • Manage conversation context by summarizing history as it approaches context limits.
Before you choose

Requirements & limits

The repository is public and specifies an MIT license. The documented installation uses pip or uv.

  • The framework is designed to let the model choose its task approach rather than impose a rigid workflow.
  • Code execution can be configured locally, in Docker or in an E2B sandbox.
  • LiteLLM, Docker, E2B, MCP and browser integrations are listed as optional installation extras.
  • The README's example uses OpenRouter; web search requires a separate Brave API key.
  • The repository is the Python implementation; StirrupJS is identified as a separate TypeScript implementation.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install Stirrup and prepare the documented quick-start

Set up the core package and credentials referenced by the README's example.

Before you start

  • Python package installation using pip or uv
  • An OpenRouter API key for the quick-start client
  • A Brave API key if the example should use web search

Steps

  1. Install the core package with the command below or the command below.

    Documented command
    pip install stirrup
    Documented command
    uv add stirrup

  2. Set OPENROUTER_API_KEY in the environment for the README's OpenRouter client example.

  3. Set BRAVE_API_KEY if web search is required; the README says the agent can still work without it, but web search will be unavailable.

Expected outcome

The core framework is installed and the credentials specified for the quick-start example are prepared.

Guide limitations

  • These steps cover installation and credential preparation, not running or validating an agent.
  • Web search depends on the Brave API key; without it, that tool is unavailable.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
ArtificialAnalysis/Stirrup · Official homepage title · retained in its source language
How we review sourcesSuggest a correction