PRO · deeper worked depth on demand
The base file is self-sufficient. When a job needs the worked depth, you load one module — a plain-text block you paste, zero install.
Seven curated modules carry the executable depth the base file points to: the survey wording, the formulas, the confusion-matrix cells, the worked derivations. You type Load PRO: <module> and paste the matching block. Identical bytes on every platform.
Paste the whole delimited block after typing Load PRO: pricing — the model reads it as curated depth.
The paste mechanic
Three moves, no install.
The only always-loaded overhead is the ~14-line PRO index inside the base file. Everything below loads on demand.
Name the module
Type Load PRO: pricing (or ai-evals, positioning…). This tells the assistant which curated block is coming.
Paste the block
Copy the matching pro/<module>.md file and paste it — delimited ===== BEGIN PRO: NAME ===== to ===== END PRO: NAME ===== so the boundary is unambiguous.
Get worked depth
Now the answer runs the actual method — survey wording, formulas, worked numbers — not a paraphrase. The block is plain text: identical bytes on every platform.
The seven modules
Load the depth the job actually needs.
Each card carries the exact Load PRO: command and its use-when, verbatim from the base file's PRO index.
Launching smaller? Load just three: Pricing, AI & Evals, and Positioning & GTM — the modules whose worked depth most often decides a real call. The other four are one paste away when you need them.
Pricing & Packaging
9 tools + checkWTP study, tiering, outcome/AI pricing, packaging redesign
AI Products & Evals
10 stagesfull eval plan, judge validation, monitoring/EDDOps, MVQ bars
Positioning & GTM
10 sectionsend-to-end Dunford derivation, sales-pitch build, launch plan
Experimentation & Growth
10 sectionsOEC design, growth-loop modeling, PLG motion sequencing
Discovery & Research
10 sectionsinterview guides, OST build-out, assumption-test design, needs→spec (House of Quality / QFD)
Metrics & North Star
10 sectionsNorth Star + input-metric tree, value-driver tree, R&D-portfolio / platform-renewal metrics, guardrail design, agent metrics
Strategy & Roadmap
the deep moduleStrategy Blocks & P2W, platform/product-family + multi-sided-platform strategy, integrate-vs-modularize & profit migration, Ten Types of Innovation, project-type mix, tech-commoditization, postponement, IfM roadmapping, Aumayr Business Plan
Actor-observability · never fake depth
The one rule that makes the layer trustworthy.
If the block isn't in context, the assistant can't see the curated depth — so it must not pretend to. This is the load-bearing contract, quoted verbatim from the base file.
Rule (actor-observability — never fake depth)
if a requested Load PRO block is not already in context, do NOT improvise curated depth. Reply with the exact path to paste and offer a clearly flagged lower-fidelity answer from the base file. State which you're giving.