Reading evaluation reports
Every evaluation produces a structured report. Here's what each piece means.
Click any idea in the Evaluated column to see its report. Three sections.
Composite + verdict
Big Fraunces number up top. Verdict badge (kill / refine / ship) next to it. A one-sentence reasoning string ("Strong hook, weak specifics — add a number or a timeline").
Axis bars
Eight horizontal bars, one per factor. Colors:
- Teal — 8+ (strong)
- Purple — 7 (decent)
- Amber — 4–6 (weak)
- Coral — 0–3 (dead)
Hover any bar to see the model's specific reasoning for that score. These tooltips are the useful part — read them rather than the composite.
Platform recommendation
Below the axes, a small block suggesting which of your enabled channels fits this idea best. This is derived from the axis profile + your app's channel mix. Rule of thumb:
- High shareability + moderate hook → X / Twitter thread
- High specificity + high payoff → LinkedIn long-form
- High emotional charge + contrarian angle → Reddit
- High distinctiveness + visual potential → YouTube Shorts
History
Every re-evaluation stacks in a history strip. Useful when you refine an idea — you can see the axes shift as you tighten the hook or add specifics. The latest evaluation is always the one used for downstream drafting.