Digital Products — Role Prompts¶
This directory holds reusable AI-agent role prompts specific to Digital Products work — workbook and spreadsheet-product creation, market research discipline for digital templates, marketplace commercialization, and the QA/documentation practices particular to selling structured Excel products.
These prompts supplement, and never replace, the venture's role routing at ../docs/ai/role-routing.md, which in turn maps onto the company-wide role definitions at ../../../docs/ai/agent-role-definitions.md. When a Digital Products-specific prompt below and a company-level role definition appear to conflict, the company-level definition governs; flag the conflict rather than silently resolving it in the venture's favor.
Available Prompts¶
| Prompt | Role Focus |
|---|---|
| product-manager.md | Owns product scope, requirements, and prioritization for a Digital Products product |
| market-researcher.md | Gathers and sources evidence for market-research.md and user-personas.md |
| business-analyst.md | Translates customer problems into requirements, use cases, and data dictionaries |
| workbook-architect.md | Designs tab architecture, named-range conventions, and component adoption |
| excel-engineer.md | Implements calculation and formula logic against calculation-specification.md |
| qa-engineer.md | Executes test plans, verifies acceptance criteria, and compatibility-tests workbooks |
| technical-writer.md | Writes in-product and companion customer-facing documentation |
| commercialization-manager.md | Develops pricing hypotheses, positioning, and evidence-backed listing claims |
| customer-feedback-analyst.md | Processes post-launch customer feedback and support signal into roadmap input |
| repository-maintainer.md | Maintains structural consistency of the Digital Products venture tree itself |
Using These Prompts¶
Each prompt frames a role, states what that role should always do, and states what
that role should always check before finishing a task. They are meant to be loaded
as role-framing context for an agent (or a human working in an agent-like way) doing
that kind of work — not as standalone task instructions. They do not authorize
skipping any requirement in ../AGENTS.md, company-level governance documents, or
venture-level standards.
This README establishes no policy of its own beyond describing this directory's contents and relationship to role routing.