Execution Policies

Configure how the system handles different types of operations. Choose a preset or customize per category.

Quick Presets

Hands Off

Maximum autonomy. The system executes all operations automatically with full rollback available. Best for experienced users who trust the system and want speed.

BalancedActive

Smart defaults. Low-risk operations run automatically; high-risk operations execute but require review. Rollback is always available. Recommended for most users.

Cautious

Safety first. Most operations pause for review before execution. Ideal for production environments or when working with sensitive data.

Locked Down

Maximum control. All high-risk operations are blocked by default. Only explicitly approved categories can execute. For regulated environments.

Per-Category Policies

CRITICAL RISK2 categories

Auth Changes

Modifying authentication or authorization

1d
Notify

Payment Integration

Changes to payment processing

1d
Notify
HIGH RISK3 categories

Schema Migration

Database schema changes and migrations

1h
Notify

Deployment

Deploying code to environments

1h
Notify

Data Deletion

Removing data from databases

1h
Notify
MEDIUM RISK5 categories

Code Modification

Editing existing source code

5m
Notify

File Deletion

Removing files from the project

5m
Notify

External API Calls

Calling third-party services

5m
Notify

Config Changes

Modifying system configuration

5m
Notify

Ontology Changes

Modifying the knowledge graph schema

5m
Notify
LOW RISK3 categories

Code Generation

Creating new code files and modules

5m
Notify

File Creation

Creating new files in the project

5m
Notify

Agent Registration

Registering new agent types

5m
Notify
Mode Reference

Auto

Execute automatically, rollback available

Execute + Review

Execute then present for review

Pause + Review

Pause before execution for approval

Blocked

Block unless explicitly overridden