From prd-builder
Run a structured discovery session to gather context before writing a PRD
How this skill is triggered — by the user, by Claude, or both
Slash command
/prd-builder:prd-discoverThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
```text
$ARGUMENTS
You MUST consider the user input before proceeding (if not empty).
Run a structured discovery session that produces a context file ready for PRD drafting. Ask targeted questions to understand the problem, users, scope, constraints, and success criteria. Save the output as a structured markdown file.
$ARGUMENTS is empty, ask the user for a short feature name and stop.prd/<feature-name>/prd-context-<feature-name>.md (kebab-case).Ask the user the following questions (batch where possible using AskUserQuestion). Do not proceed to writing until you have substantive answers for at least questions 1–3.
Follow up on any answers that are vague or incomplete. It's better to ask good questions than to draft a hollow document.
Once you have enough context, write a structured markdown file to prd/<feature-name>/prd-context-<feature-name>.md with this format:
# PRD Context: <Feature Name>
**Created:** <date>
## Problem Statement
<summarized from user answers>
## Evidence
<data, tickets, research cited by user — or "None provided">
## Competitive Landscape
<alternatives and their shortcomings — or "None identified">
## Target Users
<roles, permissions, personas>
## Scope
### In Scope
- ...
### Out of Scope
- ...
## Technical Constraints
<platform, language, infra constraints — or "None specified">
## Success Criteria & KPIs
- ...
## Detail Level
<full PRD / 1-pager / Shape Up pitch — default: full PRD>
## Additional Context
<anything else the user mentioned>
Tell the user the context file has been saved and suggest next steps:
/prd-draft prd/<feature-name>/prd-context-<feature-name>.md to generate a first draftnpx claudepluginhub sinositato/tato-claude-plugins --plugin prd-builderGenerates complete PRDs via guided feature discovery, product context integration, clarifying questions, and optional codebase analysis. Outputs user stories, Gherkin criteria, metrics, and launch plans.
Create structured Product Requirements Documents (PRDs) that connect problem, users, solution, and success criteria. Use when turning discovery notes into engineering-ready documents.
Writes AI-optimized PRDs with multi-AI orchestration, including clarification, research, structured requirements, and adversarial review. Use when scoping a new feature or product.