// FEATURED
The PolicyLayer MCP Registry: live risk records for 32,000 servers
Every MCP server, continuously watched: identity, tool surface, auth posture and a change feed you can subscribe to. Search it or license the data.
// BLOG
Insights & Updates
Technical deep-dives on AI agent security, spending controls, and the future of autonomous payments.
Browse by topic →// START HERE
What Is MCP Policy Enforcement (And Why Every Agent Needs It)
MCP policy enforcement intercepts every AI agent tool call and evaluates it against deterministic rules before execution. Here's how it works and how to set it up.
Why Prompt Guardrails Fail for AI Agent Safety (And What Works Instead)
System prompts can't enforce spending limits or prevent destructive operations. Here's why prompt guardrails fail for tool-calling AI agents and what works instead.
// THOUGHT LEADERSHIP
Robinhood let AI agents trade real money. The only safety check is optional. → AWS just made the case for deterministic policy at the MCP gateway → The NSA just made the case for a policy layer in front of MCP → Microsoft AGT Proved Deterministic Enforcement. Where the Next Problem Starts. → Runtime Governance Belongs at the Transport Layer → Anthropic's MCP Playbook Is for Builders. Defenders Need the Next Layer. → MCP Governance Is Table Stakes. What Comes Next? → Microsoft's Agent Governance Toolkit: 9 Packages, MCP-Blind → Bain Says Every Agentic Platform Needs a Policy Layer. We Built One. → X Just Shipped an MCP Server. It Exposes 131 Tools With Zero Access Control. → We Scanned Popular Open Source MCP Configs. Here's What We Found. → The Agent Control Problem Only Becomes Big in One World → 30 MCP CVEs in 60 Days. Most Fixes Are Solving the Wrong Problem. → The Academic Case for Deterministic AI Agent Enforcement → One Tool Call Away From a $10,000 AWS Bill → Will AI Ever Be Good Enough to Not Need Spending Limits? →
// FUNDAMENTALS
MCP Authentication: Securing How Agents and Servers Connect → MCP Authorization: Scoping What Agents Are Allowed to Do → MCP OAuth: Connecting Agents to Protected Servers → AI Agent Containment Starts at the Environment Layer → Tool-Result Injection: The MCP Attack System Prompts Miss → System Prompts vs. Transport Firewalls: Why System Prompts Do Not Equal Security → MCP Security: Why Prompt Guardrails Aren't Enough → Why AI Agent Policies Must Be Deterministic, Not Probabilistic → What Happens When Your AI Agent Goes Rogue → Why Your Agent Shouldn't Know Its Own Spending Limits →
// HOW-TO GUIDES
Stop Your GitHub MCP Agent From Force-Pushing to main → Blocking Outbound Exfiltration Through MCP Fetch and HTTP Tools → Cap LLM Token Spend on MCP Agents: Cost-Scaled Limits Beyond Call Counts → Namespace-Scope Your Kubernetes MCP Server From Production → Rotate MCP Credentials Across 30 Developers in One Click → Sandbox Your Shell-Exec MCP Server With Command Allowlists → Slack MCP Channel Allowlists: Stopping Agents Posting to #general → How to Safely Connect Claude Code to High-Risk Upstream MCP Servers → How to Safely Run AI Agents With Tool Access in Production → Your AI Agent Can Delete Your DNS Records → Your AI Agent Can Delete Every Container on Your Machine → Your Coding Agent Can Delete Any File on Disk → Your AI Agent Has Push Access to Every Repo → Your AI Agent Can Send Emails as You → Your AI Agent Can Run DROP TABLE on Production → Preventing Your AI Agent From Messaging #general → Secure Your Stripe MCP Server: Rate Limits and Spending Controls → One FLUSHALL Away From Losing Everything → How to Rate Limit MCP Tool Calls (With YAML Policy Examples) → How to Add Spending Controls to Any MCP Agent →