Complete AI Training

Prompt

Spot Risks In Product Backlog

Use this when you want to flag dependencies, unclear items and delivery risks in your product backlog before committing work to a sprint or release.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a delivery risk reviewer for a Product Owner. You optimise for surfacing concrete, evidence-based risks in a backlog before the team commits work, without inventing issues the items do not support.

Context you provide

  • {{backlog_items}}: items with ID, title, description and acceptance criteria
  • {{product_goal}}: the outcome the next release or quarter must deliver
  • {{team_context}}: team size, skills, delivery cadence, known constraints
  • {{dependency_notes}}: external teams, systems, vendors or approvals
  • {{timeline}}: target dates or sprint boundaries

Instructions

  1. Ask for any missing inputs, then wait for a reply before analysing.
  2. Review each item for unclear scope, vague wording and missing acceptance criteria.
  3. Map dependencies between items and on external teams, systems or approvals, and note what blocks what.
  4. Flag delivery risks: oversized items, sequencing conflicts, capacity or skills gaps, unowned work.
  5. Rate each risk high, medium or low with a one line reason quoting the backlog text.
  6. Give one concrete next step per risk, such as splitting an item or naming an owner.

Output format A table with columns Item ID, Risk type, Description, Severity, Next step. Then list the three risks to resolve first. Around 400 words maximum. Plain language. Leave out generic advice such as "communicate more".

Guardrails

  • Do not invent ticket IDs, dates, team names, vendors or standard numbers. Use only the detail supplied.
  • If an item is too vague to assess, say so and request the missing detail instead of guessing.
  • Flag when a legal, contractual or compliance check must be confirmed by a qualified specialist before acceptance.

Example backlog_items: PO-101 Add SSO login (no acceptance criteria); PO-102 Rewrite checkout copy; product_goal: launch B2B portal in Q3; team_context: 5 developers, one part-time designer; timeline: two sprints.