Settings

Control how Growth Inspector and Growth AI behave for this workspace. Workspace preferences, AI instructions, agent execution, permissions, connected accounts, and safety guardrails live here. Sign in to edit your workspace.

Growth AI behavior

Choose where Growth AI runs and how permission prompts are handled. These controls apply to Growth AI wherever you use it.

Agent

Checking local agent…Auto chooses the stronger available environment for each task.

Permissions

Skip all permissions disables Growth AI permission prompts. It does not bypass the app's safety rules.

Tone — how the AI should soundGrowth AI instructions & feedback

These workspace instructions are included with Growth AI requests and saved with your workspace.

Business facts the AI may useNever do / say (one per line)

Reply modeConfidence threshold — 75%

Replies below this confidence escalate to a human.

Connected social accounts

Wire up a real WhatsApp Business, Instagram, or other business account so employee-handled conversations flow into this inbox for real, instead of the sandbox tester. Sign in to connect accounts.

Backup contacts

Shown beside the inbox so anyone on the team knows who to call if the AI is ever unavailable. Sign in to manage contacts.

Always escalated to a human

These are hard-coded guardrails — the AI never handles them on its own.

• politics or government criticism
• religion or religious rulings
• gender / relationship sensitivities
• legal advice
• medical diagnosis or advice
• pricing or refund commitments beyond stated policy