prompt-injection-scanner

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Prompt Injection Scanner

Purpose and Intent

用途与目标

The

prompt-injection-scanner

is a security tool specifically for the AI agent era. It identifies weak points in agent instructions where a malicious user could potentially "hijack" the agent's behavior by inserting conflicting instructions into input fields.

prompt-injection-scanner

是一款专为AI Agent时代打造的安全工具。它能够识别Agent指令中的薄弱环节，这些环节可能会被恶意用户利用，通过在输入字段中插入冲突指令来“劫持”Agent的行为。

When to Use

使用场景

Skill Development: Run this every time you update the
```
capabilities
```
or instructions for an agent skill.
Pre-deployment Security Review: Essential before making an agent accessible to untrusted users.
Continuous Security Auditing: Periodically scan all skills as new injection patterns are discovered.

技能开发阶段：每当更新Agent技能的
```
capabilities
```
或指令时，都应运行此工具。
部署前安全审查：在向不可信用户开放Agent访问权限前，这一步至关重要。
持续安全审计：随着新的注入模式被发现，定期扫描所有技能。

When NOT to Use

不适用场景

Standard Code Auditing: Use the
```
secret-leak-detector
```
for credentials; this is specifically for "instruction-level" security.

标准代码审计：若需检查凭证泄露，请使用
```
secret-leak-detector
```
；本工具专门针对“指令层面”的安全问题。

Input and Output Examples

输入输出示例

Input

输入

yaml

skill_path: "./agent-skills/data-processor/SKILL.md"

yaml

skill_path: "./agent-skills/data-processor/SKILL.md"

Output

输出

A structured report highlighting parts of the instructions that are susceptible to prompt hijacking, along with concrete mitigation strategies.

一份结构化报告，突出显示指令中易受提示劫持影响的部分，并提供具体的缓解策略。

Error Conditions and Edge Cases

错误情况与边缘案例

Missing Instructions: If a skill defines tools but provides no behavioral instructions, the scanner will flag this as a risk.
Complex Logic: Highly conditional instructions can be difficult to model and may result in false positives or negatives.

缺失指令：若某个技能定义了工具但未提供行为指令，扫描器会将其标记为风险项。
复杂逻辑：高度条件化的指令可能难以建模，可能导致误报或漏报。

Security and Data-Handling Considerations

安全与数据处理注意事项

Metadata Focus: Only scans instructions; does not touch private user data.
Local Analysis: Recommended to run locally within the development environment.

聚焦元数据：仅扫描指令内容，不会触及用户私有数据。
本地分析：建议在开发环境中本地运行该工具。