Skip to main content

Browse the directory

Showing 1 resource for "evals"
Saved
Active

All 1 results are trusted trust — check install risk per entry.

Adoption queue

Browse adoption queue · balanced

0/1 visible results are ready for staged adoption under this preset.

ready 0caution 1hold 0

Agent Evals Regression Gate Skill

1 blockers: Metadata review

caution

70/100

Request metadata review from maintainers or internal owners.

skills/agent-evals-regression-gate · trust trusted · confidence 83%

Decision confidence

Decision confidence scan · balanced

1/1 results are high-confidence for the selected preset.

high 1medium 0low 0

Build repeatable eval suites that catch quality regressions in AI agent behavior before merge or release.

Level:advancedType:generalVerified:draft
Safety ✓ Privacy ✓