Back to skills

skills/prd-goal-delivery/SKILL.md

prd-goal-delivery

Drive PRD implementation through a durable /goal, scoped parallel implementation agents, independent QA ownership, acceptance evidence, and persona-based test instructions. Use when a PRD must be delivered end-to-end and every acceptance criterion must be verified before marking the goal complete.

npx skills add https://github.com/flpbalada/fb-skills --skill prd-goal-delivery
GitHub

Skill Docs

Goal

Create new /goal for this PRD:

Deliver every acceptance criterion successfully.

First Pass

  • Read PRD.
  • Extract scope.
  • Extract acceptance criteria.
  • Define evidence required for each criterion.
  • Identify target personas affected by new functionality.

Agent Setup

Spawn parallel agents for independent workstreams. Use as many agents as useful, no more.

Give each agent:

  • Dedicated /goal.
  • Clear scope.
  • Expected output.
  • Evidence required.

Required Agents

Agent Responsibility
Implementation agents Build independent pieces.
QA agent Own independent quality verification.

QA Responsibilities

QA agent must:

  • Build acceptance checklist from PRD.
  • Verify each criterion with precise evidence.
  • Run or request tests/checks.
  • Reject incomplete or unverified work.
  • Report gaps, risks, regressions.

Completion Gate

Do not mark /goal complete until:

  • All implementation results are synthesized.
  • QA confirms every acceptance criterion passes.
  • Final answer includes acceptance checklist with status and evidence.
  • Final answer includes persona-based test instructions.

Final Answer Format

Acceptance Checklist

Criterion Status Evidence
[criterion] pass / fail [command, output, screenshot, file, or behavior proof]

Persona Test Instructions

For each affected persona, include:

  • Persona name / role: [name and role]
  • Goal: [goal user wants to achieve]
  • Steps: [steps to test new functionality]
  • Expected result: [expected result per step]
  • Edge cases / failure states: [states user should try]

Failure Loop

If any criterion fails:

  • Keep /goal active.
  • Dispatch follow-up fix agent.
  • Re-run QA on failed criteria.