AI
Mode
Pick a mode and start writing.
Brand rules & guidelines
These control how the AI writes. Edits apply to every new response — no code changes needed. Core safety guardrails are always enforced on top of these.
Copywriting modes
Edit the output format for each mode, or disable modes you don't use.
Settings
Account and AI provider status.
How guardrails work
Every AI request is built on the backend as: fixed safety guardrails → your editable brand & compliance rules → the selected mode's output format → recent session context → your request. The fixed guardrails and rules are sent as a system instruction the user messages cannot override. The OpenRouter API key never leaves the server.