Protocol
Ablate a skill before you trust it
If removing the skill changes nothing, the skill may be documentation for humans rather than leverage for the model.
When it fits
- A prompt, skill or instruction bundle is believed to improve an agent.
When to avoid it
- Small eval sets can miss rare benefits; retain safety-critical instructions when removal risk cannot be responsibly tested.
Why it matters
Run representative tasks in clean environments with the skill enabled and disabled. Repeat enough trials to see whether the difference survives model variability. Compare outcome quality, trigger accuracy, cost and context use. Prune instructions that do not change useful behavior, and keep the eval so future model or harness upgrades can show when the skill is no longer needed.
Steps
- Choose representative tasks and graders.
- Run clean trials with the skill.
- Run comparable trials without it.
- Compare outcomes, trigger failures and cost.
- Prune or revise no-op instructions; keep the regression test.
An example
A code-review skill that adds 1,500 tokens but catches no additional defects across repeated tasks should be simplified or retired.
Check your result
The skill has evidence of incremental value over the same workflow without it.
Keep this limit in mind
- Small eval sets can miss rare benefits; retain safety-critical instructions when removal risk cannot be responsibly tested.
Evidence and sources
AI Engineer recommends evaluating skills on repeatable tasks, including clean runs with and without the skill, because a skill that does not change outcomes may only consume context.
Ablation results depend on task set, model and grader quality.
Don't Ship Skills Without Evals · Test cleanly, repeat, and compare skill behavior