A universe of events and platforms focused on open source, open tech and the open web.

All Things Open

Agent Identity Protocol (AIP) is an open source framework being proposed to the IETF to address the security gap in agentic AI systems. It introduces a two-layer model: Layer 1 for agent identity and Layer 2 for tool call enforcement. The core problem is that AI agents today typically receive unrestricted 'God mode' access to services like GitHub or production databases, with no policy layer governing their actions. AIP enforces least-privilege-by-default, where agents start with zero capabilities, unknown tool calls are denied, and all decisions are logged to an immutable audit trail. A hands-on tutorial walks through using the AIP Go Proxy to intercept MCP (Model Context Protocol) JSON-RPC traffic, define a YAML policy that allowlists specific tools, and demonstrate both allowed and blocked tool calls against a test echo server.

How to secure agentic AI with Agent Identity Protocol (AIP)

An open source framework for least-privilege permissions and audit trails in agentic AI.

The proble m: AI agents are running without guardrails

Tutorial: Securing an MCP filesystem server with the AIP go proxy