Assumption-first planning and design gap analysis. Use when user is in PLANNING mode, explicitly asks to plan or discuss, asks for a gap analysis / design review / stress-test of a spec or design document, or when agent faces open decisions — including headless tasks where the deliverable is proposals and open questions as files. Instead of interrogating the user, the agent proposes concrete tagged answers in small verify-style rounds (or directly in the output documents when no user is in the loop), surfaces contradictions as questions, and asks only the questions that genuinely fork the design.
78
91%
Does it follow best practices?
Impact
56%
1.05xAverage score across 1 eval scenario
Passed
No known issues
Stress-testing design gaps via awkward scenarios
Concrete weird scenarios
100%
100%
Design silence identified
100%
100%
End-to-end trace
100%
83%
Proposals are exact
100%
100%
Themed organization
0%
0%
Dependency chains noted
100%
100%
Safe-default vs judgment-call tags
0%
0%
Implied sub-decisions surfaced
87%
87%
Open questions format
0%
0%
Triage discipline
0%
25%
Contradiction surfaced as question
20%
60%
Settled decisions respected
100%
87%
No implementation deliverables
60%
40%
Table of Contents