Paste a trace, get a verdict — a tiny diagnostic page on Pages

[akash-coded]

The problem

The Four Verdicts is three
booleans and a lookup table, and people still get it wrong — usually by classifying from the
answer rather than from the trace.

The proposal

A static page on the docs site. Paste a trace JSON, get the verdict, the reasoning, and a link to
the fix. No server, no data leaving the browser — which matters, because traces contain queries.

Extend it to a list of traces and it renders the distribution, which is the thing that
actually decides where the month goes.

Why this earns a slot

The framework is already written and the logic is already implemented in
L11. This is a rendering job over code that
exists and is tested, and it makes a mental model usable by someone who has not read the docs.

The bit that needs a decision

Trace shapes differ per stack. Do we define a minimal schema and ask people to map onto it, or try
to sniff common shapes? I would define the schema and be strict, but that is the bit worth
arguing.