Setting objection handling rules
In the app, objection guidance is typed in the agent editor, not a separate “objection rules” database table.
Where it lives
Open Edit agent → Step 4 — Prompt & Goals.
Use the Objection Handling field (Markdown-supported). The in-app help says content is merged into the prompt as OBJECTION HANDLING. You can use bullets, short tables, or prose—whatever matches how your team documents rebuttals.
For stop / compliance behavior, also use Guardrails & Do's and Don'ts on the same step (merged as guardrails in the runtime prompt).
Optional Tenant platform rules packs on the Platform rules page apply before persona and tools; use them for non-negotiable policies that should not be omitted from any agent.
What the static should show: Objection Handling markdown editor with a short example list visible; character Total prompt row above or below the section.
Character budget
Objection Handling shares the same Total prompt limit as Identity, discovery, close, custom instructions, goals, and greetings (40,000 characters, per product limits). If you hit the limit, shorten other sections or goals—the UI blocks moving to the next step or saving until the total is under the cap.
What the GIF should show: Pasting a compact objection table, watching the total character badge update.
Review loop
After changing objections, use Call logs and Call details to confirm the agent stays on-message; see Reviewing agent performance.
What the video should show: Edit objection text → save → open a recent call transcript where a pushback occurred.