Free Prompt Analysis
Start with a diagnostic on your current prompt.
Two paths. Have us write you a voice-first prompt over a quick voice intake, or paste the one you already use — pulled live from Bland or VAPI if that's easier — and get an 8-category Prompt Whisperer Audit covering structure, voice fit, edge cases, and scaling risk.
What we look at
Prompt structure
Cognitive load
Flow logic
Interrupt handling
Transfer rules
Persona consistency
Latency contributors
Edge-case complexity
Redundant instructions
Prompt modularity
Scalability risks
Step 1 — Where should we send your prompt?
The intake takes about 2 minutes. You'll get the prompt by email shortly after.
Step 2 — Voice intake
Tap "Start a Call" to start the conversation
Our agent will ask a few quick questions about your business, who calls in, and what the voice agent needs to do. Speak naturally — there are no wrong answers.
When you're done, just hang up. We'll take it from there.
We're writing your prompt.
Thanks — we'll email a polished, voice-first prompt to in a few minutes.
Want a score on the result? Paste it back through the Evaluate flow once it lands.
Questions? Write to mike@prompt-whisperer.com
What's in your FREE Prompt Whisperer Audit
- Executive summary of your prompt
- Scored on 8 voice-specific criteria
- Your 2–3 biggest strengths
- Your 2–3 most critical weaknesses
- Top 3 prioritized improvements
- A rewritten example of your weakest section
Step 1 — Where's your prompt?
0 characters
Used once to fetch your prompt. Never stored.
Used once to fetch your prompt. Never stored.
Step 2 — Where should we send the audit?
Optional Context — helps us tailor the audit
Your prompt is never shared publicly. Audits are reviewed by Mike before delivery.
Sample audit
FULL AUDIT ---------- # Voice AI Prompt Audit: Bland Car Rentals Total Score: 46 / 80 Rating: Needs Significant Improvement ## 1. QUICK SUMMARY This prompt has strong foundational instincts — the Midwestern tone, backchannel logic, and customer empathy are thoughtful touches. However, the prompt is significantly over-engineered and internally contradictory, which is likely causing the agent to behave unpredictably and contributing to caller hang-ups. ## 2. SCORECARD Voice-First Language: 5/10 Persona Clarity & Consistency: 6/10 Conversational Flow & Turn-Taking: 4/10 Handling Real-World Call Behavior: 6/10 Conciseness & Signal-to-Noise: 3/10 Human-Sounding Behavior: 6/10 Edge Cases & Failure Modes: 5/10 Business Outcome Alignment: 5/10 ## 3. WHAT'S WORKING Tone and persona guidance is genuinely thoughtful — "amiable concise cheerful apologetic... midwestern charm... 'no worries, take your time!'... 'Minnesota nice,' with a sprinkle of brevity." ## 4. WHAT NEEDS IMPROVEMENT The prompt repeatedly contradicts itself: "Don't repeat yourself..." vs. "You then need to read back their Rental Record Number..." This causes decision paralysis at runtime. ## 5. TOP 3 PRIORITIZED IMPROVEMENTS 1. Simplify the core flow to: greet → ask why they're calling → request and read back the record number. 2. Fix the opening to reduce hang-ups in the first 5 seconds. 3. Replace digit-counting logic with a simple backchannel rule the model can actually execute.
You're all set.
Your full audit will be emailed to you within 1–2 business days.
⏱ Usually within 1–2 business days
Calculating your score…
— / 80
Questions? Write to mike@prompt-whisperer.com
What we received
Audit sent to
Source
Platform
Use case
Pain point
Your prompt