Messaging Directives

What are messaging directives?

Messaging Directives (at AI Assistant > Messaging Directives) are workspace-wide behavioral rules that apply to every AI assistant in your workspace, regardless of their individual persona or playbook.

Think of them as standing orders that sit above everything else. While a persona tells an individual assistant how to behave, a messaging directive tells all your assistants how to behave.

Directives vs. personas

It's worth understanding the difference:

Use messaging directives for rules you want to enforce consistently across your entire operation — things that should never vary between assistants.

Common uses

Response formatting

If you want all your assistants to communicate in a specific way, directives are the right place to set that:

• "Keep all responses under 3 sentences."
• "Never use bullet points or numbered lists. Always write in full paragraphs."
• "Always end your response with a question to keep the conversation going."

Tone and language

• "Always use formal, professional language. Never use contractions."
• "Respond in the same language the customer uses."
• "Avoid jargon and technical terms. Use plain language."

Compliance and legal

• "Never make specific claims about pricing. Always direct pricing questions to a human agent."
• "Do not provide medical, legal, or financial advice."
• "Always include this disclaimer at the end of any response about our services: [disclaimer text]."

Creating a directive

1. Go to AI Assistant > Messaging Directives.
2. Click Add Directive.
3. Write the directive as a clear, specific instruction.
4. Click Save.

The directive takes effect immediately for all new conversations. Existing active conversations will pick it up on the next AI response.

Managing your directives

Keep your directive list focused. A long list of conflicting or overlapping directives can confuse the AI and produce inconsistent results. Review your directives periodically and remove any that are no longer needed or that contradict each other.

If a directive is causing unexpected behavior in a specific assistant, check whether it conflicts with that assistant's persona. Persona instructions and directives can sometimes pull in opposite directions.