Verified65% confidenceFactExact time
The principle of indirect prompt injection is that an attacker disguises malicious instructions as normal data, delivers them to the model through a data source trusted by the agent, and induces the model to interpret the data as instructions and execute them.
3
Sources
65%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
VerifiedIndirect prompt injection refers to attack instructions hidden in external web pages, documents, or API responses accessed by an agent, where the malicious instructions are activated when the agent retrieves and injects them into the context.78% similarUnverifiedPrompt injection is a novel and unique attack vector targeting operational agents; malicious web pages can inject instructions into an agent via visual elements or hidden text, and how to distinguish real instructions from forged ones remains an open problem that has not been fully solved.75% similarUnverified攻击者通过提示词注入控制Agent的决策逻辑后，可利用Agent已获授权的所有工具和接口实现全链路攻击，形成混淆代理人（Confused Deputy）问题74% similarUnverifiedIn Agent scenarios, because model behavior is unpredictable, the importance of using short-lived credentials is higher than in traditional software systems, and the attack surface may also include attacks targeting the model itself, such as prompt injection74% similarUnverifiedIn Agent systems, a model's hallucinated output may be used to trigger tool calls, write to databases, or send notifications, causing errors to cascade and amplify across multiple steps.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/535979API
curl https://kongchang.com/api/v1/knowledge/claims/535979MCP
get_claim(id=535979)