Skip to main content
← Back to list
01Issue
FeatureShippedSwamp CLI
Assigneesstack72

Relationships

#612 Test skill evals with Fable in multi-skill eval tests

Opened by stack72 · 6/10/2026· Shipped 6/10/2026

Problem

The multi-skill eval test suite currently does not exercise skill evals against the Fable model. As Fable becomes a first-class model option, we should verify that skills behave correctly when evaluated by Fable — its response patterns and tool-use behavior may differ from Opus/Sonnet in ways that surface skill regressions.

Proposed solution

Add Fable as a model target in the multi-skill eval test runner. This would run the existing skill eval suites against Fable alongside the current model targets, catching any model-specific regressions in skill behavior.

02Bog Flow
OPENTRIAGEDIN PROGRESSSHIPPED+ 1 MOREASSIGNED+ 2 MOREREVIEW+ 3 MOREPR_MERGED+ 1 MORENOTIFICATION_SKIPPED

Shipped

6/10/2026, 6:22:33 PM

Click a lifecycle step above to view its details.

03Sludge Pulse
stack72 assigned stack726/10/2026, 4:57:25 PM

Sign in to post a ripple.