Inkog
Scans AI agents for vulnerabilities like prompt injection, tool misuse, and token bombing across 20+ frameworks, mapping findings to compliance standards.
Inkog is a security scanner for AI agents that identifies agent-specific vulnerabilities such as prompt injection, tool misuse, and token bombing across 20+ frameworks including LangChain, CrewAI, and MCP servers. It targets security leaders, DevOps, and AI engineers in enterprises and mid-market companies. The product is delivered as a SaaS platform with a CLI and MCP server, offering static analysis and adversarial testing, and maps findings to compliance frameworks like EU AI Act and NIST AI RMF.
Key features
- Static analysis for AI agents
- Adversarial testing
- MCP server integration
- Data Flow Graph analysis
- Compliance mapping to EU AI Act
- Compliance mapping to NIST AI RMF
- Shadow AI discovery
- Prompt injection defense
- Supports 20+ frameworks
- CLI and MCP server
- X<0.1/day
GTM channels
- Blog
- Marketplace
- Community
- API
- Docs
ICP
- Security teams
- DevOps sre teams
- Software developers