Press Release

Your organization's credentials are probably already on the dark web. The question is whether your security team knows which ones, which systems they...

When an AI agent decides to call a tool, the default behavior in most MCP-compatible frameworks is simple: the call goes through. No policy check, no...

Every tool call an AI agent makes is a decision point. Without a gate at that point, the agent's judgment is the only control in place. vaikora-guard-mcp...

Pull the package from PyPI, point it at your MCP server, and every tool call your agent makes will hit a policy check before it executes. That is what...

Phishing campaigns fail when the infrastructure supporting them is identified during staging, before any email is sent. Cyren's detection approach is...

Most AI agent frameworks let tool calls through the moment the LLM decides to make one. There is no gate between "the model chose this action" and "the...

Open-source security tooling earns trust the only way that actually works: anyone can read the code, run it locally, and verify what it does before...

Security tools protect what they can identify. Access controls restrict what they know exists. DLP policies block what they can categorize. If a file,...

Twelve lines of Python and a policy file. That is the minimum viable guardrail you can put between an AI agent and any tool it calls, using the...