This is the manual version. Start Pocket Agent if you want the system to run for you.
Discovery → MVP Prompt Pack
Discovery call Monday. Working MVP by Friday. Same prompts, every time.
Discovery call to working software is where most operators stall.
You took the notes. You have a transcript. You have a sense of what the buyer wants. Then you sit down to build and realize you don't have a spec — you have impressions. Three sessions in, the agent is shipping things you didn't ask for and missing things you did.
Patrick was different. Discovery call on Monday. Working roof-inspection software live the following Friday. $3,500 build, monthly hosting, signed handoff. Same agent everyone else uses.
The difference is what got handed to the agent between the call and the first build prompt. That's what this pack is — the prompts that did the handing.
Eight sequenced prompts and the case study they shipped.
Eight prompts in order, each one fed by the output of the last. Run them on a real discovery transcript and you walk out with a spec, a feature list, dispatch-ready build prompts, branding extracted from the client's existing site, and a handoff doc. The Patrick case study is the proof — every prompt run end to end, real outputs, real result.
- 1The Discovery Call Framework — 9 questions to ask in a 30-minute call, the feature-parking-lot technique, red flags that mean this isn't a build (it's a rescue), opening / closing scripts
- 2Transcript → Spec prompt — paste the transcript + one paragraph of framing, get a structured spec: problem, audience, MVP feature set, post-MVP backlog, open questions
- 3Spec → Feature-List prompt — break the spec into shippable features ranked by leverage (impact × buildability), with dependencies and estimated dispatch lanes
- 4Feature → Build-Prompts sequence — each feature becomes a dispatch lane prompt with verbatim success criteria, blockers, and verification rules
- 5Branding Extraction prompt — pull logo, colors, type, and copy tone from the client's existing site; generate Tailwind config + brand tokens. Plus the no-brand-yet path (defaults + 3 options the client picks from)
- 6Migration Path prompt — for clients with existing CSVs, sheets, or current-SaaS exports. Generates extract → transform → load + reconciliation checklist
- 7Handoff Doc prompt — generates the 60-day support handoff (how to log in, make changes, add users, read analytics, who to call when X breaks)
- 8Patrick Fresh Page Home Improvement — the full case study. Transcript → MVP arc, every prompt with the actual output, annotated with what worked and what I'd do differently next time
PDF + markdown prompt files + the annotated case study. The prompts are the sequence. Patrick is the proof that the sequence works against a real client, real dollars, real signed handoff.
The rest of the stack.
Each kit is $15. Same shape, same instant download, same refund-on-reply guarantee.
- [ $15 · instant download ]The Dispatch Playbook
The field manual for running parallel Claude Code agents without them stepping on each other.
Get it for $15 - [ $15 · instant download ]The Dev-Team Document Set
Eleven document templates plus three operational conventions — drop them into any repo and the agents inherit dev-team discipline on day one.
Get it for $15 - [ $15 · instant download ]The CLAUDE.md Template Library
Six opinionated CLAUDE.md templates (SaaS, contractor SaaS, marketing site, mobile app, API service, internal tool) — pick the closest, fill five slots, agent is on-rails in thirty minutes.
Get it for $15 - [ $15 · instant download ]Wire the Brain to Your Stack
Seven MCP walkthroughs — Drive, Gmail, Slack, Notion, Linear, GitHub, Supabase — with auth setup, working queries, real gotchas, and the brain-sync patterns operators actually run.
Get it for $15
Want all five? Bundle for $60 — save $15 against $75 buying them separately.
Five $15 playbooks = $75. All five bundled together — Dispatch, Dev-Team Document Set, CLAUDE.md Template Library, Discovery → MVP Prompt Pack, Wire-the-Brain-to-Stack — at $60.
Your Pocket Agent gets sharper every week.
This playbook is one piece. Pocket Agent captures everything your AI needs and gives back — standup before coffee, brief before every call, customer Q&A in your voice. Pocket Agent gets sharper every week. You don't lift a finger.
Tap once. Your AI captures the rest.
Voice memos, screenshots, share links, emails, Loom recordings — all filed automatically. Nothing dies in your head.
Your AI works while you sleep.
Standup before coffee. Brief before every call. Customer Q&A in your voice. Content from past wins. It just runs.
Want this running without the setup?
The playbook teaches the pattern. Pocket Agent runs it — every playbook active from day one. Sign up and you're running.
You can buy this kit and build the workflow yourself.
Or start Pocket Agent and have your AI run the workflow for you.
[ $97/mo · 14-day free trial · Skool community included ]
$15. Instant download. Close one $3,500 build and the pack paid back 230×.
$15. Instant download. The prompts hit your inbox the moment Stripe confirms — eight markdown files, the PDF guide, the Patrick case study end to end.
If the sequence doesn't get you from a real discovery transcript to a real working build, reply and I'll refund it. No form, no question.