Teams treat a certificate—or a “full catalog” purchase—as implementation maturity. Then they discover checkout sold the foundation path, not every module in the production bundle. That gap is why promptanatomy.app exists as a conversion hub, not as proof that the workflow is owned.
This field note records what the hub sells, how it differs from the blog and marketing site, and what counts as practice versus a governed workflow. Source and deploy details live in the open mother repo DITreneris/promptanatomy; the interactive course ships as a git submodule from DITreneris/inzinerija. If you are mapping properties for the first time, start with The Prompt Anatomy Ecosystem Map.
What promptanatomy.app is
promptanatomy.app is Platform HQ—not the knowledge hub and not the discover-and-try demo.
promptanatomy.blog (this site) publishes free frameworks, governance playbooks, templates, and case studies. Read here when you need RACI worksheets, eval gates, or audit trail patterns.
promptanatomy.site (launched 2025-04-12) is discover-and-try: ecosystem journey map, five-part Anatomizer, 60-second maturity quiz. See Prompt Anatomy Marketing Site Launch.
promptanatomy.app is where reading becomes repeatable drills: the 6-block prompt system, knowledge checks, business scenarios, context engineering, and a capstone. Stripe checkout sells the foundation path (modules 1–6). Later paths are operator grants. Conversion and plans and pricing live on the hub root—not on deep training routes.
The hub fills the gap between “I understand the framework on the blog” and “my team executes the same steps under feedback.”
Three “Anatomies” readers confuse
Prompt Anatomy uses “anatomy” at three layers. Mixing them causes teams to buy training when they need governance—or paste Anatomizer exports into production without owners.
| Layer | Where | What it teaches |
|---|---|---|
| 6-block prompt system | Training app (post-purchase, /anatomy/ path) |
How to write structured prompts—Meta, Input, Output, Reasoning, Quality, Advanced |
| 5-part Anatomizer | .site |
Persona, Context, Variables, Instructions, Constraints—discover and demo |
| 6-layer implementation stack | .blog Foundations |
Outcome → workflow → context → model → eval → governance—how teams operate |
All three are intentional: write prompts in training, discover structure on .site, operate workflows on .blog. For the founder essay on why an operating system beats a chatbox, see Beyond the Chatbox on Medium.
What checkout sells
The production training bundle holds modules 1–12. Stripe Phase 1 does not sell that whole bundle.
| Path | Modules | How you get it |
|---|---|---|
| Starter (39 EUR) | 1–3 | Stripe checkout |
| Core (99 EUR) | 1–6 | Stripe checkout |
| Data Analysis | 7–9 | Operator grant — not checkout |
| Agent path | 10–12 | Operator grant; tester cohort by email — see 3A Before You Build an Agent |
| B2B workshop (399 EUR) | Live session, 3 hours, ≤10 people, license excluded | Not a Stripe plan |
Core is the paid foundation path, not the full bundle. Starter and Core are one-time, lifetime access—no subscription expiry. See plans and pricing on the hub.
Access is email-based: checkout through Stripe, then magic-link entry from the Pricing section—no separate password account. From this blog, start at the hub root or pricing section, not a gated deep link.
Progress persists locally; the app includes glossary, tools catalog, and a prompt library with copy actions. Community support runs through the public Telegram group.
Hub stack and deploy shape
| Surface | URL / repo |
|---|---|
| Hub | promptanatomy.app |
| Training SPA | /anatomy/ on same domain (post-purchase) |
| Mother repo | github.com/DITreneris/promptanatomy |
| Training submodule | github.com/DITreneris/inzinerija |
Hub frontend: Vite + React, Tailwind, LT/EN locale routing (/, /en, /lt), SEO head per route.
Commerce backend: Production checkout runs on Vercel api/; FastAPI is local and CI. Stripe webhooks upsert Supabase user_access (highest plan per purchaser). Magic-link entry follows from the Pricing section.
Deploy: Single Vercel project—landing at root, training SPA under /anatomy/, legacy /anatomija/ 301. CI “Golden Legacy” builds frontend + submodule with production VITE_* parity and runs backend pytest.
Intentional thin scope: No CMS on the hub. Training content ships as versioned JSON (modules.json, glossary.json, promptLibrary.json)—content edits do not require React refactors.
From checkout to owned workflow
The launch hero encodes the hub chain: Plans → checkout → magic link → module progress → governed workflow on .blog.
Random chat produces enthusiastic individuals. Structured blocks produce designed artifacts. Training produces repeatable drills. The blog produces owned workflows with eval gates and RACI. Sponsors fund the last step—not the first demo. The hub is not the runtime—see The Model Is Not the System. Checkout and module progress live on .app; governed workflows live on .blog.
That progression mirrors how we write on .blog: diagnose with 10 Signs Your Company Is Vibe Prompting, design on the AI Workflow Canvas, gate with Evaluation Hooks for AI Workflows.
Launch guardrails
Treat .app completion as practice and purchase, not proof of enterprise implementation maturity:
- A certificate or finished Module 3 does not replace a documented workflow ID, RACI, or eval gate pack on your side.
- Do not paste hub or training copy into procurement decks; link the relevant playbook on
.blogand cite pass rate, cycle time, or incident cost on real workflows. - Free ecosystem spokes (
.cloud,.info,.space,.help,.ceo,.pro,.lol) complement daily work; they do not replace the paid foundation path on.app. For the Enter spoke—first AI lesson and quick send check—see Quick Send Check First; for the Use spoke—five-locale daily org-context library—see Who Orchestrates the Builders; for the Manage spoke—weekly CEO brief pattern and AI Operations Center—see The Weekly CEO Brief Pattern; for the Play spoke—optional brand game on Telegram—see Corporate Ladder Soft Launch.
We learned this on sister properties: Classroom Prompt Builder had a webhook pointed at .app while buyers returned to .online—payment succeeded, fulfillment failed. Same-host checkout, webhook, and success URL is a launch guardrail, not an implementation detail.
promptanatomy.app gives practitioners a structured front door into training. The job of this blog remains turning that practice into repeatable, owned AI workflows—with owners, eval gates, and audit trails that survive the next model swap.
