All products

Review product decisions with AI

Clarify the problem, intended outcome and trade-offs before adding another feature to the roadmap

Review product decisions with AI by tracing a proposed feature back to the problem it should solve. The pm-thinking skill asks who experiences the difficulty, how often it occurs and what happens if nothing changes. It gives product discussions a basis beyond a list of things someone would like to build.

Use it when examining requirements, goals, priorities or a roadmap. It separates shipping an output from achieving an outcome and encourages small releases that can teach the team something. The method supplies questions and challenges, rather than a guaranteed answer. Your team retains judgement over the evidence, constraints and trade-offs.

What it does

  • State the problem before choosing a solution
  • Define an outcome with a baseline
  • Compare alternatives and explicit non-goals
  • Record the reasoning behind a decision
  • Separate current work from later possibilities
  • Identify a smaller useful experiment

Install and use

  1. Make the skill and its published product-philosophy context available through the documented folder layout. Use a file-capable agent and prepare the decision, evidence and team constraints; no private workplace tool is required.
  2. For example: “We are choosing between better documentation search and a new dashboard. Use these observations to clarify each option’s problem and success measure, and ask about missing evidence.”
  3. Answer questions about the affected users, frequency and intended change. Keep the baseline distinct from a target and mark assumptions that remain untested.
  4. Review what each choice gives up. Keep the decision and next experiment in the project’s notes. This review does not itself authorise implementation or publication.

FAQ

Source

Made by Rian van der Merwe. Original skill: pm-thinking, published on GitHub under the MIT licence.

View on GitHub