Agents and harnesses

Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection

by Kai Greshake et al.

Read the original →

Foundational public research on indirect prompt injection. Linked only; no payloads reproduced.

The paper studies risks when language-model applications encounter instructions through external content. This entry links to public security research for threat modeling and intentionally omits payloads and exploit procedures.

Level

Expert

Models

any

Category

Agents and harnesses

Tags

security, prompt-injection, research

Source date

2023-02-23

Added

2026-10-18

Last verified

2026-10-18

Status

ok

Summary in our own words. All rights with the original creator. Correction or removal: DM on X.