此内容尚未提供您的语言版本,正在以英文显示。
Demo Metrics
命令行 已验证 活跃Open-source AI agent firewall for MCP security: agent egress control, DLP, SSRF, and prompt injection defense.
目的
To secure AI agent workflows by acting as a firewall at the network boundary, preventing data exfiltration, prompt injection, and other security threats.
功能
- 11-layer URL scanning pipeline
- Process sandbox with OS-native primitives
- Bidirectional MCP scanning
- Signed action receipts for attestation
- DLP and prompt injection detection
使用场景
- Securing API keys and credentials from AI agent exfiltration
- Preventing prompt injection attacks on AI models
- Auditing and attesting to AI agent actions
- Integrating AI agents into regulated or high-security environments
非目标
- Replacing the AI agent itself
- Providing AI agent development tools beyond security
- Acting as a general-purpose network proxy without security focus
质量评分
已验证99 /100
13 days ago 分析