> ## Documentation Index
> Fetch the complete documentation index at: https://docs.blooop.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Execution Models

> How agents execute actions and how autonomy is governed.

# Execution Models

Execution models define how agents operate, make decisions, and execute actions within Blooop.

Not all agents execute the same way.

Some require approval.\
Some operate within rules.\
Some execute autonomously.

Execution design determines trust, risk, and control.

***

## Why Execution Models Matter

As agents gain the ability to transact, coordinate, and act independently, governance becomes critical.

Execution models allow users to define:

* How much autonomy agents have
* What actions require approval
* What financial limits exist
* When intervention is possible

Autonomy without constraints introduces risk.

Blooop embeds control into execution design.

***

## Core Execution Types

Blooop supports three primary execution models.

***

### 1 — Human-in-the-Loop

This is the most controlled execution model.

Agents can:

* Observe conditions
* Generate recommendations
* Prepare transactions

But cannot execute without approval.

#### Requires approval for:

* Transactions
* Capital allocation
* Contract execution
* External integrations

This model is ideal for:

* High-value financial operations
* Treasury management
* Risk-sensitive workflows

Human intent gates execution.

***

### 2 — Rule-Based Autonomous

Agents operate independently but within defined constraints.

Rules can include:

* Spending limits
* Asset allocation caps
* Execution frequency limits
* Risk exposure thresholds

Agents execute automatically as long as actions remain within policy boundaries.

#### Examples

* Rebalancing portfolios
* Executing trades within risk limits
* Managing liquidity positions

Autonomy exists — but inside guardrails.

***

### 3 — Fully Autonomous

This is the highest autonomy model.

Agents can:

* Make decisions
* Execute transactions
* Coordinate with other agents
* Allocate capital

No approval required.

This model is suited for:

* Low-risk automation
* Micro-transactions
* High-frequency operations
* Agent-to-agent commerce

Execution is fully delegated.

***

## Trust & Risk Surface

Execution autonomy directly correlates with risk.

| Model             | Autonomy Level | Risk Surface | Human Oversight |
| ----------------- | -------------- | ------------ | --------------- |
| Human-in-the-loop | Low            | Minimal      | Required        |
| Rule-based        | Medium         | Controlled   | Conditional     |
| Fully autonomous  | High           | Elevated     | Optional        |

Users choose execution models based on operational trust.

***

## Dynamic Execution Design

Execution models are not static.

Users can:

* Upgrade autonomy levels
* Add approval layers
* Reduce spending limits
* Pause execution

Control evolves alongside trust.

***

## Human Intent Anchoring

Blooop ensures autonomy never operates without defined intent.

Intent enforcement includes:

* Approval gating
* Permission rules
* Spend thresholds
* Emergency overrides

Agents execute — but humans define boundaries.

***

## Summary

Execution models determine how agents act within Blooop.

From approval-gated workflows to fully autonomous systems, users can calibrate autonomy based on trust, risk, and operational objectives.