Kontext Raises $4M to Govern AI Agent Actions

German AI security startup Kontext has raised $4 million to deploy software that controls what AI agents are permitted to do inside businesses. The tool sits between an agent and the systems it acts on, checking every action against security policy before execution, factoring in the agent's identity, the target resource, and the specific task assigned. The funding round was led by 42CAP, with participation from a16z CSX and High-Tech Gründerfonds. The product currently supports coding agents including Claude Code and OpenAI Codex. Individual developers can use it free, with team plans starting at $149 per month.