OUTCOMES
What this Prompt helps you deliver
- An industry domain model and workflow map exist
- TabTin reuse and new capability boundaries are clear
BEFORE YOU START
Before you start
- A real business process is already operating
- You can describe actors, inputs, outputs, and exceptions
What AI should inspect first
docs/principlepackages/appsapps/tabtin_djangoapps/tabtin-electronREADY TO RUN
Complete Prompt
Copy the complete Prompt with its execution and acceptance boundaries.
Model a real industry process into a TabTin-based product by understanding how it creates value before choosing UI or tables. Read AGENTS.md, docs/principle, packages/apps, apps/tabtin_django, and apps/tabtin-electron. Interview me about customer, payer, intake, actors, inputs, decisions, deliverables, compliance, exceptions, and human fallback. Use our business language.
Map acquisition/intake through execution, review, delivery, repeat business, and exception paths. Extract stable domain objects, lifecycle, owner, auditable events, and permission boundaries. Map them to Project, Task, Workspace, Agent, TabData, TabDoc, and App; add an object only when existing semantics do not fit. Mark human, Agent, collaborative, and mandatory-approval steps. Choose the first deliverable closed loop rather than building the whole industry platform.
Replay three real cases: normal, missing information, and handoff/human intervention. Every state needs entry conditions, owner, visible output, and next step, and customer value must be measurable. Have a nontechnical operator read the summary. Report glossary, relationships, states, TabTin mapping, v1 scope, extensions, and open decisions. Do not begin large implementation until a closed-loop plan is approved.