EvalForge: Agent Evaluation Lab | Codelit.io

Skip to content

You're offline. Some features may not work until you reconnect

EvalForge: Agent Evaluation Lab

3a · 4t · 3s · 1m

Drag to reorder. Use Option-Up, Option-Down, or More actions.

EvalForge: Agent Evaluation Lab

Catch behavior regressions before agent workflows reach production users.

Dry run ready

Starts from on

or

starts from on

1

Uses

produces

from

2

Uses

produces

from

3

Uses

produces

from

For

Workflow changed to Scenario Designer. Handoff: Starts.
Nightly suite to Scenario Designer. Handoff: Starts.
Scenario Designer to Team output. Handoff: On failure.
Scenario Designer to Eval Runner. Handoff: Always next.
Eval Runner to Team output. Handoff: On failure.
Eval Runner to Release Judge. Handoff: Always next.
Release Judge to Team output. Handoff: On failure.
Release Judge to Team output. Handoff: Produces.