# Gather Requirements

> Use when a delivery run needs PRD-style requirements produced from a user request: extracting every atomic requirement, rule, decision, and story into a classified requirement register, naming actors and protected actions, recording assumptions, open questions, and acceptance criteria. This is the gather_requirements action of the planning stretch — one Agent-tool dispatch with subagent_type business-analyst — distinct from resolve-open-questions (which answers the questions this action raises) and promote-to-product-knowledge (which files the resolved rows).

- Skill: `bhangeef16/gather-requirements` (Agent Skill)
- Install (CLI): `npx skillmds@latest add bhangeef16/gather-requirements`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bhangeef16/gather-requirements/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: BhangeeF16 (https://skillmd.com/u/bhangeef16)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/bhangeef16/gather-requirements

---


# Gather Requirements

Produce the PRD-style scope for exactly one resolved target. This action extracts and classifies; it does not answer its own open questions, design anything technical, or promote anything to canonical knowledge — those are later actions.

## Dispatch

One synchronous `Agent`-tool call, `subagent_type: business-analyst`. The persona file carries the domain lens (questions to challenge, review rules, stop conditions); this skill defines the action's own contract.

## Required inputs

- The user request.
- Relevant feature or capability docs for the target.

## Required outputs (all mandatory — do not skip one because it seems redundant)

- `planning.md` for the target.
- A requirement register: every atomic requirement/rule/decision/story extracted from the request, classified by canonical destination, each row carrying a status of draft/questioned/resolved/promoted.
- Assumptions, recorded explicitly.
- Open questions, recorded explicitly.
- Acceptance criteria naming actor, action, expected outcome, and failure path.

## Security checks (required within this action)

- Identify actors and protected actions.
- Identify tenant and resource scope.
- Identify abuse cases for sensitive workflows.

## Rules

- A requirement that conflicts with an existing business rule or decision, with the conflict unresolved, is a stop condition — surface the conflict as an open question rather than silently picking a side.
- If the target turns out to be larger than one run can meaningfully cover, stop and ask how to split it — never silently narrow scope.

## Result

Report each required output produced (with its location), the result (pass / pass_with_risks / blocked / failure), and blocking questions as an explicit list (empty list stated explicitly).