Blog

Product news and best practices for teams by Runlayer.

Protecting yourself from last week’s prompt injection attacks

Two near-invisible prompt injection attacks showed how attackers can bypass default enterprise guardrails and trigger silent, ongoing data exfiltration by exploiting user and model trust. Runlayer blocks these attacks by treating every input as untrusted until it passes continuously updated security models trained on the latest real-world exploits.

January 19, 2026

AlL posts

Runlayer Product Update - 1.25.0

This update is about momentum: moving faster in the CLI, getting clearer visibility into what’s running, and debugging with less friction. Expect smoother workflows, better control, and fewer surprises as you build and ship.

Announcing Cursor and Runlayer’s partnership on Cursor Hooks

Runlayer is an official Cursor Hooks launch partner. With Cursor Hooks, securely allow or deny MCP tool calls with Runlayer's enterprise MCP platform.

Marcin Jan Puhacz

Runlayer Joins Anthropic, OpenAI, & Google as AAIF Founding Member

The Linux Foundation has launched the Agentic Artificial Intelligence Foundation (AAIF), with Runlayer joining sponsors Anthropic, OpenAI, Google, AWS, Microsoft. AAIF now oversees the Model Context Protocol (MCP), reinforcing MCP as a rising standard for AI agent integration. Runlayer supports AAIF’s open, secure, and scalable AI development mission.

Your AI agent is probably leaking your data right now

MCP adoption is accelerating across major platforms, but security risks—like malicious servers, prompt injection, and tool-level exploits—are growing just as fast. This post breaks down real attack scenarios that show how easily data can leak when MCP implementations are trusted by default. It also outlines practical defenses for users and builders, plus why companies need audited MCP catalogs, gateway proxies, and sandboxing to stay secure at scale.

Runlayer Raises $11M to Scale Enterprise MCP Infrastructure

Why MCP builders are transitioning from DCR to OAuth CIMD

Over the last year, MCP has surged in adoption. To little surprise, this has introduced some scaling issues. One of these is client registration; previously, systems were rigged together by humans. Today, AI agents discover and interface with MCP servers freely, requiring a new paradigm for client communications.

What is Dynamic Client Registration?

Tired of manually registering every AI agent with every OAuth server? Dynamic Client Registration (DCR) lets your agents authenticate with MCP servers at runtime, no human clicks required. Learn how DCR works, when to use it over traditional OAuth, and why it's becoming essential for scalable agentic systems.