CI/CD On-Call AI Agent: Architecture, Permissions, and Safeguards
An agent can accelerate the collection of evidence during an incident, provided that it starts in read-only mode, separates diagnosis and action, and measures its reliability.
Topic
MCP
MCP, A2A and AGNTCY: understanding the interoperability of AI agents in 2026
Not all agentic protocols solve the same problem. Stacking them without a clear need increases the attack surface and operational complexity.
MCP in business in 2026: architecture, authorization, governance, and migration to the 2026-07-28 specification
MCP standardizes the connection between AI applications, data, and tools. It does not standardize trust, business rights, or governance: these responsibilities remain to be designed.