This page requires JavaScript to display.
The Nolan Scott Team Sells and buys businesses from $1M to $15M+, and advises on the real estate that comes with them. nolanscottteam.com →
Example 18 / 18 · a sample from our Rolodex of 100+ automations Claude Skill
Automation Breakdown · Claude Skill

Coyle Evaluator

Turns raw mystery-shopping field notes into finished luxury-hotel and restaurant evaluation reports: notes reconciled against the folio and receipts, chronological narratives written to exacting format rules, expense rows generated, and the 200-field ShopMetrics survey filled by the browser, leaving only the final submit to the human.

200+ survey fields auto-filled
every mismatch flagged first
1 assumptions log per report
0 auto-submits
True AI automation isn’t one clever workflow. It takes 20 to 40 workflows working together to make a real impact on a business. That’s what we do: a Rolodex of 100+ automations across different CRMs and platforms, ready to be custom-built for your specific CRM and software stack.
The Workflow
Scroll sideways · click any step for its explanation
Reconciliation finds the findings
A management letter only became a real finding because a cross-check noticed it. Real discrepancies surface when notes meet folio, not before.
The assumptions log is a deliverable
Every inferred answer is tracked and surfaced for sign-off, not produced only when asked.
Never click Submit
The skill fills, saves, and verifies. The final submit always belongs to the evaluator.
▶
New Evaluation
hotel or restaurant
F
Upload Guidelines
client objectives PDF
⇄
Split Reports?
hotel vs F&B
F
Field Notes In
raw + conversational
{ }
Reconciliation Pass
notes vs folio
⇄
Mismatches?
evaluator decides
A\
Write Narratives
section by section
{ }
Assumptions Log
every inference
{ }
Reimbursement Rows
tab-separated
Ch
Auto-Fill ShopMetrics
JS injection
Ch
Save + Verify
scroll, don’t read back
⇄
Evaluator Sign-Off
assumptions confirmed
◈
The report at a glance
Click any step to see exactly what it does. The pipeline reads left to right: guidelines and reconciliation, format-exact narratives with an assumptions log, then a browser-driven survey fill that stops short of Submit.
Claude in Chrome
Documents in
Claude writing
Logic / rules
Decision gate
Trigger
How It Works

Reconcile, write, fill, never submit

01 · Guidelines + Reconcile
Every client is different

Each brand has its own objectives PDF, required outlets, and quirks, nothing carries over between properties. Before a word is written, the field notes are cross-checked against the folio, receipts, and property emails: room numbers that don’t match, impossible charges, misremembered menu items. Every mismatch goes back to the evaluator with both values.

02 · Write to Spec
Narratives with rules

Chronological play-by-play with timestamps and staff quotes, a separate paragraph per dish, strict staff-description formats, no em dashes, no procedural filler, a thoughtful guest recounting the stay, not a robot. Every answer inferred rather than stated in the notes lands in the assumptions log for sign-off.

03 · Auto-Fill
200 fields, zero submits

The browser fills the ShopMetrics survey section by section, narratives pasted, radios and time dropdowns set, every No answer given its comment, using the form’s own internals rather than pixel-clicking 200 fields. Expense reimbursement rows come out ready to paste. The Submit button is never touched.

Next example: LinkedIn Autopilot →
AI · nolan scott team
Unpacking...