Prompt-Ops Maturity Check — Free Diagnostic

Eight questions, about three minutes. If your team runs on AI prompts and workflow instructions — in chat tools, docs, or scripts — this scores how safely that prompt capital is managed, and how close your best workflow is to being a reusable, licensable asset. Everything is computed in your browser. No account, no tracking, nothing leaves this page.

This free diagnostic runs on Final Odyssey, the tools and diagnostics site of Castle Creative — the same operation behind the fixed-price Workflow Automation Audit at flowaudit.castlecr.com. Diagnostics live here; the paid offer and delivery live there.

1. Where do your production prompts and workflow instructions live today?

A versioned, searchable library (repo or tool) the team actually uses Shared docs or a wiki page — findable, but not versioned per change Chat histories, personal notes, tribal knowledge Mostly in people's heads, rewritten on the fly

2. When you improve a prompt, what happens to the old version?

Kept; the change is reviewed; rollback is one step Kept somewhere, but rollback means manual archaeology Overwritten — the old wording is gone We don't deliberately change prompts

3. Before a changed prompt goes back into use, how do you know it still works?

A fixed set of test inputs and expected checks runs against it One or two manual spot checks We notice later, when output degrades We don't check; quality drifts

4. Who owns the prompt library?

A named owner, plus contributors with clear review One person, informally Everyone and no one Haven't thought about it

5. How often does someone rewrite a prompt that already existed somewhere?

Rarely — the library is the first stop Sometimes, when they can't find it Often — duplicate prompts are normal here Constantly — every use is a fresh rewrite

6. When a prompt or workflow quietly breaks or degrades, how fast do you notice?

Monitoring or alerts catch it near real-time A user or customer tells us within days We notice after it has hurt output for a while We generally don't notice

7. Do you know what your AI workflows cost per task or per output?

Yes — tracked per workflow A rough total from invoices No breakdown — a gut feeling No visibility at all

8. Could a competent third party run your best workflow today, without you?

Yes — documented and packaged: inputs, outputs, limits, failure modes defined Mostly, with some hand-holding Only with the author in the loop No — it is inseparable from our heads Roughly how many AI workflow uses does your team run per week? (optional, sharpens one finding) Minutes typically spent writing or tuning one prompt? (optional)

–

Findings

Do this first

Licensing readiness

If this diagnostic found the work: Final Odyssey is Castle Creative's tools site — the fixed-price named-workflow audit and its delivery live at flowaudit.castlecr.com. Scoped remediation after a free proof, never payment-before-proof. Email trinity.williams@castlecr.com and paste your result (button above). A person reads it. No auto-responder, no list, no invoice on this page.

Final Odyssey free diagnostic — a Castle Creative tools site (the paid audit offer lives at flowaudit.castlecr.com). This page is self-contained: no external scripts, styles, fonts, or analytics. Nothing is sent off this page. Refresh clears everything. No invoice on this page.