Skip to main content

Exit Criteria: the Gate Review

Gate 2 → 3: Passage conditions
  • The Solution Brief is complete with all eight sections and has been signed or explicitly approved by all relevant internal stakeholders.
  • The solution has been validated with clients/users and they confirm it solves their problem.
  • The Decision Matrix is documented with weighted criteria fixed before evaluation.
  • The Assumptions Register is complete. High-risk assumptions have been verified or have a documented contingency plan.
  • The Feedback Log is closed: all received feedback has been incorporated or discarded with justification.
  • The Translatability Map has been reviewed and there are no dimensions with insufficient definition for Context Phase.
  • The Corridor Consensus Test has been passed: each stakeholder can explain what will be built and why, consistently.

Do not advance if:

  • Approval is only hierarchical and there is no real understanding of the solution by stakeholders.
  • There are unresolved and undocumented disagreements.
  • The Solution Brief has unverified high-risk assumptions without a contingency plan.

After the gate: transition to Phase 3​

When the team completes the Gate Review, something important happens: the team gains the right to translate its thinking into AI-processable language.

This concept — earning the right to context engineer — is central to the methodology. A team that has achieved real consensus arrives at Phase 3 (Context Phase) with the precision needed to produce context that actually works. A team that skipped alignment arrives with ambiguity that contaminates every prompt, every agent configuration, every output.

The Solution Brief travels to Phase 3 as the direct input, accompanied by the Problem Statement, the Translatability Map, and the full set of artifacts. Together they contain:

  • The solution defined — what will be built, for whom, and what it changes.
  • The reasoning behind it — why this solution and not the alternatives, with explicit criteria.
  • The risks that accompany it — assumptions classified by impact, with action plans.
  • The translation readiness — how precisely each dimension is defined for AI context.

Phase 3 doesn't start from scratch. The structuring work done in this phase is the raw material for Context Phase:

  • The Solution Tree becomes the structure of the PRD that guides the PM Agent.
  • Active assumptions become constraints in the context that the AI must respect.
  • Success criteria become the metrics that AI agents must optimize.
  • The Translatability Map becomes a quality checklist — dimensions flagged as risk signals require more iteration during context design.

There is an element that deserves special attention in this transition: the Problem Statement travels intact. It is not rewritten. It is the canonical reference. The Solution Brief, however, is translated, not copied. The Context Engineer transforms it into processable documents: the PRD for the PM Agent, the Architecture Document for the Architect Agent, the Story Files for the SM Agent. The four-dimension structure enormously facilitates this translation because each dimension maps to a specific aspect of the context.