review-trace

name	review-trace
description	Walk a concrete scenario end-to-end through the architecture — find gaps and friction

Scenario Trace Review

Take a concrete scenario (user story, data flow, or command sequence) and trace it through every component it touches. This catches gaps between "the design says X" and "but then how does Y actually happen?"

ultrathink

Rules

• You must trace a SPECIFIC scenario, not review in the abstract. If no scenario is given, pick the most representative user workflow from the design.
• At each step, name: the component, the interface used, the data shape, and what could go wrong
• "Works fine" is not useful. Show the actual path and flag where the design is silent or hand-wavy about the transition

Invocation

/review-trace [target] [scenario]

If no scenario specified, choose the most common/critical user workflow described in the design.

Workflow

1. Identify the scenario: What is the user or system trying to accomplish?
2. Trace entry: Where does the scenario enter the system? What triggers it?
3. Follow each hop: Component → interface → next component. At each transition:
   • What interface is used? (API call, context update, event, file read)
   • What's the data shape at this point?
   • What could go wrong at this transition?
   • Is this transition documented or implicit?
4. Trace exit: Where does the scenario produce its final result? What does the user see?
5. Flag gaps: Where did you have to guess because the design doesn't specify?
6. Flag friction: Where does the scenario require unnecessary hops or transformations?

Output Format

Scenario: [description]

Step 1: [User does X]
  → Component: [name, § ref]
  → Interface: [what's called/triggered]
  → Data: [shape at this point]
  → Risk: [what could go wrong]
  → Gap: [anything unspecified]

Step 2: [Component handles X, passes to Y]
  → ...

Summary:
- Gaps found: [list]
- Friction points: [list]
- Unspecified transitions: [list]