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

Autoharness

Autoharness is an open-source Python framework that adds governance and supporting infrastructure around AI agents, including context and token-budget management, tool-call validation, permission checks, risk classification, output sanitisation, cost attribution, session persistence, multi-agent profiles and JSONL audit logging.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Autoharness — 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 Autoharness when an AI agent needs explicit controls around tool execution, context growth, cost visibility, prompt-injection handling or audit records rather than a direct ungoverned model integration.

Who it is for

  • Developers building AI agents that call external tools
  • Teams needing configurable governance around agent runs
  • Engineers evaluating self-hosted Python agent infrastructure

Documented uses

  • Wrap an OpenAI client with a governance layer
  • Control and inspect agent tool calls
  • Manage context budgets and truncation
  • Attribute model costs per call
  • Record agent decisions and provenance in JSONL audit logs
Before you choose

Requirements & limits

The repository is public, active and MIT licensed. The documented setup uses Python 3.10 or newer and local installation with pip.

  • The framework is Python-based and documented for self-hosted use.
  • Its governance pipeline includes six named stages from validation through audit logging.
  • Core, Standard and Enhanced modes trade pipeline depth and hooks against overhead.
  • The repository documents OpenAI client wrapping and a separate AgentLoop interface.
  • The evidence describes an Enhanced mode warning and identifies it as the default mode.

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

Set up Autoharness around an OpenAI client

Install the repository and add its documented wrapper around an OpenAI client.

Before you start

  • Python 3.10 or newer
  • An OpenAI client integration
  • A local clone of the public AutoHarness repository

Steps

  1. Clone the AutoHarness repository.

  2. Change into the project directory and install it with pip install -e .

  3. Import OpenAI and AutoHarness, then wrap the OpenAI client with AutoHarness.wrap(OpenAI()).

Expected outcome

The OpenAI client is wrapped by AutoHarness so the documented governance layer can be used around agent requests.

Guide limitations

  • This procedure documents setup and client wrapping; it does not establish that a particular model, tool configuration or deployment will operate successfully.
  • The repository example uses an OpenAI client and does not document provider compatibility beyond that example.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
aiming-lab/AutoHarness · Official homepage title · retained in its source language
How we review sourcesSuggest a correction