What Does 'Skill Poisoning' Mean?

[ SLANG ][ AI CULTURE ][ AGENT SECURITY ]

FORCE: Machine CultureUNIVERSE: Agent / Developer SecurityCLUSTER: Agent Attack Vectors

An attack where malicious instructions are embedded inside an AI agent's installed skill or tool definition itself, causing the agent to behave harmfully when that skill is invoked, without the agent's underlying model being compromised directly.

REAL-WORLD EXAMPLE

"Security review caught skill poisoning in a third-party skill, buried instructions that would've exfiltrated data on every call."

LORE & ORIGIN

Emerged in AI agent and developer security discourse as agent skill ecosystems grew, naming this specific attack vector on the skill layer rather than the model or infrastructure layer.

FIRST SEEN

2026-06

PEAK POPULARITY

2026 (Agent / Developer Security)

CURRENT STATUS

Mainstream Gen Alpha Slang

阅读中文版 →日本語版を読む →한국어판 보기 →Leer en español →Ler em português →Lire en français →

EQUIVALENT CONCEPTS IN OTHER LANGUAGES

中文 zh-CNSkill Poisoning →日本語 ja-JPスキルポイズニング →한국어 ko-KR스킬 포이즈닝 →Español es-ESskill poisoning →Português pt-BRskill poisoning →Français fr-FRskill poisoning →

AURA IMPACT ⓘ

+5 AURA

TREND STATUS

↑ HIGH RISING FAST · HIGH

CULTURE CATEGORY

[ SLANG ][ AI CULTURE ][ AGENT SECURITY ]

RELATED TERMS

MENTIONS OVER TIME

2023202420252026

TOP PLATFORMS

TikTok

68%

YouTube

15%

Twitter / X

Others

WHEN DID YOU FIRST HEAR THIS?

CLICK AN OPTION BELOW TO CAST YOUR VOTE.

[ VOTE TO REVEAL COMMUNITY RESULTS ]