Skip to main content

Agents

Prompting

Writing instructions that survive a real phone call.

What a prompt is in Finn

A Finn's behavior comes almost entirely from its prompt. The Behavior tab splits that prompt into three fields (Identity Text, Style & Guardrails, Response Guidelines), plus the Welcome Message that plays on connect. Everything on this page is about what to put in those fields.

Write the prompt in four parts, in this order. Skip one and the agent gets brittle on live calls.

PartWhere it livesLength
IdentityIdentity Text1-3 lines
GoalIdentity Text or Response Guidelines1 line
RulesStyle & Guardrails5-15 bullets
ExamplesResponse Guidelines2-4 micro-dialogues

Identity

Anchor a name, a role, an employer, and a tone. Nothing else.

You are Aria, a friendly appointment coordinator at Smile Dental Clinic.
You speak with patients calling to confirm, reschedule, or cancel an upcoming visit.
You are warm, brief, and never push extra services.

Backstory costs latency and gives the model room to improvise. Three lines is the ceiling.

Goal

One sentence, one concrete outcome.

Goal: confirm, reschedule, or cancel the patient's upcoming appointment within 90 seconds.

A goal like "help the caller with whatever they need" makes the agent meander and calls run long. If the real job is multi-step (qualify, then book, then warm-transfer), split it into separate nodes in the Workflow editor instead of stacking it into one prompt.

Rules

State positives and negatives. The "never" rules matter as much as the "always" rules, because they are the only thing standing between the model and an improvised answer it should not give.

Rules:
- Always ask "Am I speaking with {customer_name}?" before sharing any appointment details.
- If the caller says "remove me" or "stop calling", honor immediately and end the call politely.
- If asked if you're a real person, say honestly: "I'm Aria, a voice assistant for Smile Dental."
- Never quote a price — transfer to the front desk if asked.
- Never schedule a Sunday appointment — the clinic is closed.
- If the caller's intent is unclear after two clarifying questions, offer human transfer.

Rules that are only negative leave the agent with no positive frame, and it will stall rather than recover. Pair each prohibition with what to do instead.

Examples

Two to four short dialogues anchor tone better than any adjective. Show edge cases, not the happy path you already described in the rules.

Example 1 — Smooth confirm
Aria: "Hi, is this Priya? It's Aria from Smile Dental, calling about your visit on Thursday."
Caller: "Yes."
Aria: "Great. We have you booked for 4 PM Thursday with Dr. Mehta for a cleaning. Still works?"
Caller: "Yes, that works."
Aria: "Confirmed. See you Thursday. Have a great day."

Example 2 — Reschedule
Caller: "I need to move it."
Aria: "No problem. We have openings Friday at 11 AM or 3 PM, or Monday at 10. What works best?"

Examples that restate the rules add nothing. Write the angry caller, the ambiguous request, the caller who answers a different question than the one asked.

Variables

Use {variable} syntax inline. Finn injects audience columns and live call data at runtime.

VariableResolves to
{customer_name}The audience row's name column
{customer_phone}The dialed number
{current_time}Caller's local time, formatted
{call_duration}Seconds since pickup

Any column in an uploaded audience CSV becomes a variable, so {appointment_date} and {loan_amount} work if those columns exist. See variables.

Length and latency

LengthLatencyFits
Under 300 wordsLowestConfirm, reminder, survey
300-800 wordsNormalMost agents
800-1500 wordsAcceptableComplex sales, triage
Over 1500 wordsSlow and erraticSplit into a workflow

Past 1500 words the agent gets slow and starts contradicting its own rules. That is the signal to move logic into the workflow editor or out into a knowledge base, which holds facts the prompt should not carry inline.

Common failures

MistakeWhat goes wrongFix
Multi-paragraph identityBloats the first token, slows responseThree lines
Vague tone ("be helpful")No effect on outputThree concrete adjectives
Open-ended goalAgent meanders, calls run longOne outcome per agent
Negative-only rulesAgent stalls with no fallbackMix positive and negative
No examplesTone drifts under pressureTwo to four
Examples mirroring the rulesRedundantShow edge cases

Debugging a prompt

When calls go off-script, do this in order:

  1. Listen to the recording in call logs. Do not guess from the transcript summary.
  2. Find the first turn that goes wrong. Every turn after it is downstream noise.
  3. Add one rule covering that exact scenario, plus one example.
  4. Re-test on the same number using Test on Call from the Finn page.

If step 4 does not fix it, the problem is not the prompt. Look at the workflow, the audience data, or the knowledge base. Most "the model is dumb" reports are a missing rule for a scenario nobody wrote down.

Prompt edits are made in the dashboard on the Finn's Behavior tab, not through an API field. Changes apply to new calls only. In-flight calls finish under the configuration they started with.