Clockworke × NuStrength — Project Proposal
CLOCKWORKE Digital
Project Proposal No. CWD-NS-2604
Automate Scale · Success
● ● ●
For NuStrength ● ● ● Prepared April 2026

Functional
Lab Reports,
on autopilot.

A Claude-native reporting system for the RYM Method coaching team — replacing a 7-step manual workflow with a single tool, a single review loop, and a sub-three-minute turnaround per client report.

Prepared for
NuStrength
Attn: Craig McDonald
Prepared by
Ann Camille Balon
Founder · Clockworke Digital
Date issued
April 2026
Valid 30 days
Scope
Phase 01
clockworke.com
S / 01 Executive Summary
The situation in one page

World-class clinical work,
trapped in a seven-step workflow.

NuStrength's Functional Lab Reports are deeply researched, clinically nuanced, and beautifully branded. The production process, however, is unsustainable — a manual workflow stitching ChatGPT and Claude together, consuming roughly an hour of senior time per report, with zero headroom as the client base grows.

Clockworke Digital proposes a lean, Claude-native system that replaces every step of that workflow with a single tool, a single intelligence layer, and a single review point.

The coach submits client data once — screenshots, blood panel PDF, notes — and receives a complete, branded, client-ready report in under three minutes. No GPT. No template-stitching. No server infrastructure to maintain. Just Claude, configured correctly.

Investment at a glance

One-time build USD $2,500 (two milestones of $1,250).  ·  Optional maintenance USD $80/hr, on-demand, no retainer.  ·  Delivery 3 weeks from kick-off.  ·  Bonus 2 × Skill Refinement Sessions within first 60 days of go-live.

One-time build
$2,500 USD
2 milestones
Time saved / report
~55 min
60 → 5 minutes
Annual value
~468 hrs
at 10 reports/wk
Go-live
3 weeks
from kick-off
S / 02 Understanding the Brief
From the briefing call & sample reports

The bottleneck isn't the analysis —
it's the production line.

We studied the briefing call and the Chelsea Brice and Valerie Lippe sample reports. The picture is clear, and so is the fix.

  1. The clinical intelligence is already world-class

    Coaches produce nuanced, archetype-driven reports — Digestive Slowdown, Metabolic Suppression, Hormonal Instability — with pro-metabolic framing, functional range interpretation, and RYM Method-aligned recommendations.

  2. The bottleneck is the workflow, not the thinking

    Screenshots are manually interpreted, data is pasted into ChatGPT (hitting an 8,000-character instruction ceiling that forces constant prompt editing), output is copied into a Canvas, formatted as a PDF draft, reviewed, then re-dropped into Claude for branding before final distribution.

  3. One hour of senior time per report. Ten-plus reports a week.

    Ten hours of expert time lost weekly to a task that should take five minutes — and the ceiling hardens as the client base grows.

  4. The future-state vision is bigger than a tool

    A full dashboard and GHL integration where coaches generate reports directly from the NuStrength Tracker app data. This proposal addresses the current bottleneck in Phase 1 and positions the automation roadmap as a Phase 2 conversation — scoped once the core system is live and the time savings are measurable.

S / 03 The Architecture Decision
The most important page in this proposal

Two philosophies.
One of them compounds.

The way this system is built determines whether it scales, evolves, and compounds — or whether it becomes another piece of infrastructure to maintain. Two approaches look similar on the surface and diverge dramatically in the real world.

A
Approach A

The hybrid stitch

Keep the two-tool architecture. GPT does clinical analysis. Claude does formatting. A hardcoded PDF template. n8n orchestration. A reference-range database to keep GPT from hallucinating. Validation layers. A hallucination-mitigation framework. This is how most automation agencies think.

B
Approach B · Recommended

Claude-native

One tool. Encode the full clinical intelligence — functional ranges, RYM Method principles, brand voice, report structure, archetype logic — into a Skill. Claude reads screenshots and PDFs natively. Generates the branded HTML report in a single step. Coach reviews. Download PDF. Done.

Side-by-side: what changes for NuStrength

Dimension
Hybrid stitch
Claude-native
AI architecture
GPT + Claude — two tools, two APIs, two points of failure
Claude only — one tool, one conversation, one review loop
Hallucination risk
Requires a 5-layer mitigation framework because GPT tends to invent reference ranges
Reference ranges live inside the Skill — grounded source, nothing to hallucinate
Infrastructure
Self-hosted n8n automation server that NuStrength has to maintain
Native Claude Projects — no server, no maintenance
Report output
Hardcoded PDF template — rigid, same shape for every client
Dynamic HTML artifact — adapts to archetype
Coach workflow
Submit data, wait for pipeline to run through multiple APIs, review output
Drop everything in one message — report back in under 3 minutes
Evolution
Each change = developer work + redeployment
Coaches can improve the Skill themselves by updating reference files
Ongoing cost
Claude API + GPT API + n8n hosting + maintenance
Claude API only — $10–20/month at current volume
Why this matters in plain terms

The hybrid stitch is rebuilding the existing broken workflow with better plumbing — automating the symptoms. The Claude-native approach fixes the root problem: the process is fragmented because the tools were fragmented. Put it all in one place and the whole thing collapses from seven steps to three.

S / 04 Proposed Solution
One project, one Skill, one download button

A dedicated Claude Project
with a custom Skill.

We will build NuStrength a dedicated Claude Project with a custom Skill that encodes the full intelligence of the current report production system. The Skill replaces the ChatGPT prompt, the template formatting, the brand application, and the clinical interpretation framework — all in one place.

What lives inside the Skill

  • 01

    The complete clinical interpretation framework — functional ranges for every marker NuStrength tests, RYM Method nutrition principles, metabolic classification logic, archetype identification (Digestive Slowdown, Metabolic Suppression, Hormonal Instability, and others as they emerge).

  • 02

    The full report template — exact structure matching current production reports, rendered as branded HTML that downloads as PDF with one click.

  • 03

    Brand voice guardrails — the warm, non-clinical, pro-metabolic language that is distinctly NuStrength.

  • 04

    Unit detection — automatic metric (kg, cm, °C) vs. imperial (lbs, in, °F) rendering based on client locale.

  • 05

    Adaptive sections — Coach Cue Block, Clinical Pattern Snapshot, symptom tags, phased recovery strategies all render based on the data provided, not a fixed template.

How a coach uses it

01 · SUBMIT

Drop everything in

Open the NuStrength Claude Project. Paste client info. Drop in screenshots from the Tracker app. Attach the blood test PDF.

02 · GENERATE

Claude does its thing

Claude reads everything natively, applies the clinical framework, and generates the full branded report as an interactive artifact.

03 · REVIEW

Iterate inline

Review in-chat. "Make the supplement section gentler." "Emphasise the cycle finding." Claude revises on the spot.

04 · DELIVER

Click download

One-click PDF export. Report is ready for the client. No re-formatting. No second pass.

This is not theoretical

As part of this proposal, we've already built and tested the Skill against two of NuStrength's actual reports — Chelsea Brice and Valerie Lippe. Both generated cleanly, on-brand, with correct clinical interpretation including the Coach Cue Block format. Live demos available on request. Other proposals promise delivery; we can show delivery before you sign.

S / 05 Deliverables
Handover-ready, end-to-end

A complete system —
not just a working tool.

Documentation and training your team needs to own the system, not just use it. Everything below is included in the project fee.

01NuStrength Claude Project
Shared project space configured for the coaching team with correct access permissions.
02Custom Claude Skill
Full system prompt + knowledge base files encoding the clinical framework, brand voice, and report template.
03Functional ranges reference
Complete marker database — every test NuStrength runs, with both lab and functional ranges, embedded in the Skill.
04Archetype templates
Digestive Slowdown, Metabolic Suppression, Hormonal Instability framings encoded as Skill patterns.
05Brand-matched HTML / PDF output
Pixel-accurate replication of the NuStrength report design with brand colours, Poppins typography, and logo.
06Unit auto-detection
Metric or imperial units rendered automatically based on client data.
073 live test reports
End-to-end generation of real (anonymised) client reports with coach iteration.
08Coach SOP document
Step-by-step guide for data submission, output review, and edge case escalation.
09Team training call
60-minute live walkthrough with recording for future coach onboarding.
1030-day post-launch support
Included — answer questions, minor refinements, no hourly charge in the first month.
S / 06 Timeline
Three weeks, visible progress each phase

From kick-off
to live system.

Every phase has a tangible deliverable so you see progress continuously — not only at the end.

01Week
Discovery & Skill Build
Review existing ChatGPT prompts, sample reports, and all report variants. Build the Skill: system prompt, reference files, HTML template, archetype logic.
Deliverables
Claude Project live · Skill v1 deployed · first test report generated
02Week
Test & Refine
Run 3 end-to-end reports using real anonymised client data. Coach reviews each. Iterate on clinical interpretation, tone, and template edge cases.
Deliverables
3 refined test reports · updated Skill config · edge-case documentation
03Week
Handover & Training
Finalise SOP. Conduct team training call. Deliver recorded walkthrough. System goes live.
Deliverables
SOP document · training recording · signed-off project · 30-day support begins

Total engagement — three weeks from project kick-off.

S / 07 Investment
Transparent, milestone-based

One-time build,
split across two milestones.

Milestone 01 · end of Week 01
Discovery & Skill build
Claude Project setup · Skill engineering · template build · reference-file integration
$1,250USD
Milestone 02 · end of Week 03
Testing, refinement & handover
3 live test reports · SOP documentation · team training call · 30-day support begins
$1,250USD
Total project investment
All deliverables, fully handed over
USD$2,500
Bonus included — no additional cost

2 × Skill Refinement Sessions. Two 60-minute working sessions within the first 60 days of go-live — use them however you need: adding new archetype variants, tuning edge cases surfaced during real-world coach use, encoding seasonal protocol updates, or onboarding a new coach to the Skill. These sessions would normally bill at $160 — included free.

Ongoing maintenance — optional, no retainer

After the 30-day included support window, ongoing refinements and updates are available on-demand. Billed only for time used. No minimum, no retainer.

Maintenance typically covers
  • — New archetype patterns as the clinical framework evolves
  • — Updates to functional ranges when the reference database shifts
  • — New report sections or modifications to existing ones
  • — Expanding the Skill for new inputs (GI Map, CGM data)
  • — Troubleshooting edge cases surfaced in real-world use
  • — Training new coaches on the system
On-demand rate
$80 / hr
Billed only for time used

Payment terms
50% deposit on kick-off
50% on final delivery

Monthly operating cost
Claude API only — ~$10–20 / mo at 40–60 reports. No server hosting.
★ Return on investment

Current state — ~10 expert hours per week on report production. At a conservative $50/hr internal cost, that's $500/week × 52 weeks = $26,000/year of expert time spent on a task that should take minutes.

Future state — ~50 minutes per week reviewing Claude-generated drafts. Time saved: ~9 hours per week. This project pays for itself within the first 5 weeks of use.

S / 08 The Future Roadmap · Phase 02
Scoped after Phase 01 goes live

Phase 01 is the intelligence layer.
Every tier after builds on it.

This proposal scopes Phase 01 — the Claude Skill and the coaching team workflow. Phase 02 deliverables are not priced here; we recommend scoping them only after the Skill has 30+ days of usage data.

TIER 01 · NOW

Shared Claude Project + Skill

Coach submits data, Claude generates branded report in the Project.

~5min coach time / report
TIER 02 · AFTER 30 DAYS

Structured intake + email delivery

Tally or Typeform submission flows through Claude API. Report delivered to coach's inbox automatically.

~2min coach time / report
TIER 03 · ONCE T2 STABLE

GHL integration · full auto

Client uploads blood test to NuStrength Tracker → report auto-generates and stores on the GHL client card.

~1min coach time / report

Each tier builds on the previous one without rework. The Claude Skill built in Phase 01 is the intelligence layer that powers all three tiers — you never have to rebuild the clinical framework.

S / 09 Why Clockworke Digital
We are not a generalist software agency
We build AI infrastructure for coaches, consultants, educators and course creators. One vertical. One focus.

Clockworke Digital is a 13-person team built around one vertical — infopreneur businesses. We don't build restaurant websites or enterprise SaaS platforms. We build the marketing, funnel, and AI infrastructure that coaching and course creator businesses need to scale.

For NuStrength specifically, that means we come to this project already speaking the language — Kajabi, coaching programs, client avatars, the pro-metabolic philosophy, the Kitty Blomfield brand voice. We spent time on nustrength.com.au before writing this proposal. We read Chelsea's report end-to-end. We built the Skill against Valerie's data as a sanity check. That work is already done.

What this means in practice
  • The Skill already understands the RYM Method — we're not learning on NuStrength's dime.

  • Brand voice is calibrated to the warm, non-clinical, pro-metabolic tone that is distinctly Kitty's.

  • The functional range reference database is already drafted, ready for coach review.

  • GHL AI Employee · Claude Skills · Claude API · Artifacts — this is our core practice, not a side service.

  • We build with handover in mind. SOPs, training recordings, documentation are standard — not an afterthought.

Why not a generalist automation agency?

Most automation agencies will approach this as a generic pipeline problem — data goes in, data comes out, stitch the APIs, bill the hours. What they miss: this isn't really an automation project. It's a clinical intelligence encoding project. The hard part isn't moving data around. The hard part is making sure the system understands what a low waking temperature means in the context of a woman with stubborn fat loss, elevated monocytes, and irregular cycles.

That is what Clockworke is built to do for the coaching industry.

S / 10 Next Steps
Four actions · three weeks to live

Once the deposit is in,
we are three weeks to live.

01
Review this proposal. Optional — live demo of the working Skill using Chelsea or Valerie's data.
Owner · NuStrength
02
Share access to: existing ChatGPT prompt instructions, 3–5 completed sample reports, the NuStrength blood marker list.
Owner · NuStrength
03
30-minute kick-off call with the coaching team to walk through the current workflow and confirm archetype variants.
Owner · Both parties
04
50% deposit ($1,250) received — project formally commences. Three weeks to live.
Owner · NuStrength
Ann Camille Balon
Founder  ·  Clockworke Digital Inc.  ·  clockworke.com
Automate Scale · Success
● ● ●
Tweaks