Review Recommendation Boundary

A recommendation boundary defines what should not be recommended yet and why. It helps separate discovery findings from solution design.

A recommendation boundary is a line around what the evidence supports. It names which solutions are premature and what must be learned, checked, or decided before recommending them.

Recommendation boundaries keep the assessment grounded. They also protect clients from shiny-object recommendations that may not fit the actual work, data, risk, or ownership needs.

Do not recommend automated customer scheduling until the team confirms that technician availability data is accurate and current.

A recommendation boundary is not a refusal to help or a permanent ban on a solution. It states the conditions that must be met before a responsible recommendation can be made.

Name what must be true before you recommend.

* Premature Recommendation * Root-Cause Hypothesis * AI Candidate * Current State

* [NIST AI Risk Management Framework, "AI RMF Core"](https://airc.nist.gov/airmf-resources/airmf/5-sec-core/) — mapping context and measuring risk before managing or deploying an AI system. * [ASQ, "What is Problem Solving?"](https://asq.org/quality-resources/problem-solving) — defining and diagnosing a problem before selecting and implementing a solution.

* [[premature-recommendation|Premature Recommendation]] * [[root-cause-hypothesis|Root-Cause Hypothesis]] * [[ai-candidate|AI Candidate]]