agent-eval
ExperimentsCompareEvals

Overview

Playground for your agent evals to view experiments, evals, and compare runs.

Experiments
4
Total Runs
4
Eval Fixtures
36
Latest Pass Rate
94%

Recent Experiments

View all →
cc-mcp-opus-5.5-medium
93%1 run
cc-plugin-haiku-5.5-low
94%1 run
codex-mcp-gpt-6.1-sol-medium
86%1 run
codex-plugin-gpt-6.1-sol-low
100%1 run

Eval Fixtures

View all →
701-new-ui-flow

Build an **Order History** page that helps users look back at what they've ordered - Build the **Order History List** component listing every order t...

702-rework-ui-flow

There are several places in MealDrop where you can order dish quantities: when you browse dishes on a restaurant, and in your basket before you check ...

703-fix-bug-flow

On checkout, pressing Enter after filling in your details doesn't do anything. You have to click the button with the mouse to check out. Fix it so En...

704-fix-a11y-flow

We've had complaints from customers with disabilities that checkout is confusing to get through: after finishing the first set of fields, they can't t...

705-migrate-to-ds-flow

This app's UI is entirely hand-rolled. The company has since created a design system, and its packages are already installed in this repo. Move the ap...

706-new-ui-scheduled-flow

We want to let users schedule an order ahead of time, so they wait less for their food. - In the Sidebar cart, add a button to "Schedule delivery" ne...

Compare

Compare two experiment runs side-by-side to see pass rate deltas, duration changes, and per-eval breakdowns.

Open Compare →