---
title: "Unused skills"
description: "antiburn flags sessions that loaded a skill into context but never invoked it."
url: "https://antiburn.ai/docs/findings/unused-skills/"
---
## What it detects

A skill loaded into a session - installed, project-level, or otherwise - adds its description to every request's context so the model knows it's available, whether or not the session ends up using it. antiburn checks, for each session with real assistant work in it, whether every loaded skill was invoked at least once, and flags the session as soon as one sits unused for the whole session.

Like Unused MCP Servers, this is an absence claim, so it only fires from complete evidence. A session where antiburn's read of the invocation or context-loading records is incomplete never produces this finding, since a missed record could be the one call that used the skill.

## Why it burns tokens

A skill's description occupies context in every request for as long as it's loaded, regardless of whether the session ever calls it. An unused skill is context you paid to send on every turn without any of it doing work.

## How to fix it

Only keep the skills a given project or session actually needs loaded, and remove or disable the rest. Where skills live and how you scope them - globally, per project, or per session - depends on your agent, so check its own skills configuration for how to unload ones you don't need there.

## When it shows "not assessed"

Today this only assesses on agents whose logs attribute an invocation back to the skill that loaded it - among antiburn's supported agents, that's Claude Code. Agents that don't capture this attribution can't be assessed for this finding at all.

Even on a supported agent, a session shows not assessed rather than clean when antiburn's read of the session is incomplete - a partial tool-invocation log or a partial record of which skills were loaded - because an absence claim needs complete evidence to be trustworthy. Sessions with no real assistant work in them (idle sessions) aren't counted either way.

See [hygiene score and findings](/docs/concepts/hygiene-score-and-findings/) for how this fits into your overall score.