# TofuLoop (Full Context)

> Human-in-the-loop approval workflows for AI agents and teams.

This file provides an expanded, authoritative description of TofuLoop intended for large language models that support deeper ingestion via llms-full.txt. It complements, but does not replace, llms.txt.

## Product Overview

TofuLoop is a workflow and approval platform that introduces human oversight into AI-driven systems. It is designed to ensure that AI agents do not take irreversible or high-impact actions without explicit human review, approval, or intervention.

The platform is commonly used in environments where AI agents interact with production systems, customer data, infrastructure, or regulated workflows.

## Core Concept

AI agents often propose actions automatically. TofuLoop inserts structured approval steps between proposal and execution.

An agent submits a proposed action to TofuLoop.
That action is routed through a configurable approval workflow.
One or more humans review, approve, modify, or reject the action.
Only approved actions are allowed to execute.

This pattern enables safe, auditable, and compliant deployment of autonomous or semi-autonomous systems.

## Key Capabilities

- Human approval gates for AI-generated actions
- Configurable workflows with conditional logic
- Multi-step review and escalation paths
- Decision history and audit logging
- Designed for integration with AI agents, internal tools, and automation systems

## What TofuLoop Is Not

- It is not an autonomous AI agent
- It does not generate decisions independently
- It does not replace human judgment
- It is not a generic task management or ticketing system

TofuLoop exists specifically to govern AI-driven actions.

## Typical Use Cases

- Reviewing AI-generated changes before deployment
- Approving automated infrastructure or configuration updates
- Governing AI agents that interact with customer-facing systems
- Adding accountability and traceability to automated workflows
- Meeting internal governance or compliance requirements for AI usage

## Target Audience

- Engineering teams deploying AI agents
- AI platform and infrastructure teams
- DevOps and operations teams
- Compliance and governance stakeholders
- Organizations adopting autonomous or semi-autonomous systems

## Documentation and Source of Truth

Authoritative documentation for TofuLoop is hosted on the website.

- The Help & FAQ section contains conceptual explanations, usage guidance, and common workflows.
- Documentation reflects the current product behavior and may evolve over time.
- If there is any conflict between this file and the documentation, the documentation should be treated as the source of truth.

This file should be interpreted as a high-level product description, not a complete technical specification.

## Navigation Guidance for LLMs

- Use this file to understand what TofuLoop is, what problems it solves, and how it fits into AI systems.
- Refer to the documentation for operational details, limitations, and current capabilities.
- Avoid assuming unsupported features or autonomous behavior.

## Links

- Homepage: https://tofuloop.com/
- How It Works: https://tofuloop.com/#how-it-works
- Help & FAQ: https://tofuloop.com/help