Guardrails define what the agent will and won't do. The three core guardrails are brand voice, claim restrictions, and suppression lists. Configure all three before going Live.
Brand voice
Settings → Brand Voice. Provide 3–5 sample emails in your voice plus tone descriptors (direct, warm, technical). Every draft is checked against this profile — drafts that miss get regenerated or held.
Claim restrictions
List the claims the agent must never make. Every draft is checked against this restricted list before it sends. Common restrictions:
- Pricing promises
- Competitor disparagement
- Unverified stats
- Compliance claims
Suppression lists
Define the emails, domains, and companies the agent must never contact:
- Current customers
- Active opportunities
- Partners
- Competitors
- Opt-outs
Suppression syncs from your CRM automatically and can be extended manually.
Testing guardrails
Use Draft Preview mode to generate 10 sample drafts against real accounts before going Live. Review them for voice and claims, then adjust and regenerate until the drafts read like your best rep wrote them.
