AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

prd

Gated PRD generation — interrogates problem, user, and metric before drafting; refuses to draft on unknowns. Usage: /prd <feature-or-problem>

Original source
alirezarezvani
License
MIT
Pinned revision
e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

Ask your assistant

Use Agentic Memory skills to read alirezarezvani--prd at revision e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 1

SKILL.md

/prd

Generate a concise, evidence-gated product requirements document for $ARGUMENTS.

Usage

/prd <feature-or-problem>

$ARGUMENTS is the feature, initiative, or problem statement. If empty, ask for it before doing anything else.

Phase 1 — Forcing Questions (before any drafting)

Walk these one at a time. Do not batch them. Each answer feeds a required PRD section.

  1. Problem — What user problem does this solve, and how do you know it exists? (Evidence: support tickets, interview quotes, funnel data — "the CEO wants it" is not evidence.)
  2. User — Who specifically has this problem? (Segment, role, frequency of pain. "Everyone" is a non-answer.)
  3. Metric — What single number moves if this works, by how much, measured where?
  4. Alternatives — What do these users do today instead? Why is that not good enough?
  5. Non-goals — What adjacent asks are explicitly out of scope for v1?

Drafting Gate (hard refusal)

Refuse to draft the PRD if the answer to question 1 (problem), 2 (user), or 3 (metric) is unknown, circular, or "we'll figure it out later." Instead, output the open questions and the cheapest way to answer each (e.g., 5 customer interviews, a funnel query, a fake-door test). A PRD without a problem, a user, and a metric is a feature wish, not a requirements document.

Phase 2 — Draft (required-sections checklist)

Every PRD must contain all of these sections — emit the checklist at the end and mark each:

  • [ ] Problem statement (with the evidence from Q1)
  • [ ] Target user and segment (from Q2)
  • [ ] Goals and explicit non-goals (from Q5)
  • [ ] User stories with acceptance criteria
  • [ ] Success metric + threshold + measurement source (from Q3)
  • [ ] Alternatives considered / "do nothing" baseline (from Q4)
  • [ ] Scope, dependencies, and timeline assumptions
  • [ ] Open questions and risks

Keep it to ~2 pages. Use the repo template as the skeleton.

Phase 3 — Prioritization hook (optional)

If the user has multiple candidate features, offer to RICE-score them before committing the PRD:

python3 product-team/skills/product-manager-toolkit/scripts/rice_prioritizer.py features.csv --capacity 20

Repo Assets (verified paths)

  • Skill: product-team/skills/product-manager-toolkit/SKILL.md
  • PRD template: product-team/skills/product-manager-toolkit/assets/prd_template.md
  • PRD patterns reference: product-team/skills/product-manager-toolkit/references/prd_templates.md
  • RICE tool: product-team/skills/product-manager-toolkit/scripts/rice_prioritizer.py

Related

  • /code-to-prd — reverse-engineer a PRD from an existing codebase
  • /rice — standalone RICE prioritization