skill by

Checkpoint Promotion

Gate fine-tuned checkpoints with drift budgets, paired comparison, and forgetting checks before promotion. Use after a training run produces a checkpoint, when deciding whether a tuned model ships, or when a promoted model needs re-gating against updated goldens.

Add skill

Follow @wshobson

Permissions

Files

SKILL.md

checkpoint-promotion

Checkpoint Promotion

The Phase 5 gate for the whole plugin: a checkpoint that trains cleanly and beats its task metric still doesn't ship without clearing all four stages below. eval-harness-first built the suite re-run here — this skill is where that suite's baseline decides something.

Input: a trained checkpoint, eval/baseline-<model>.json from eval-harness-first, and the frozen eval/drift-suite.yaml. Output format: promotion-report.md — the four-stage evidence plus a terminal PROMOTE or REJECT verdict that /finetune Phase 5 and /promote-checkpoint consume directly.

2 files · 21 KB8 KB

Install

It’s free, and every skill you add syncs into every AI tool on your computer, instantly.

Get the Skillet appor

npx skilletmd add @wshobson/checkpoint-promotion -y

Version history

Drift (pts)	Verdict
≤1	Noise — proceed
2–5	Rerun with seed variation before deciding
>5	HARD FAIL — no exception for task gains

Checkpoint Promotion

Checkpoint Promotion

The Four-Stage Gate

Drift Budget

Catastrophic Forgetting

The Verdict

Related Skills