Software developers are becoming agent builders

Angie Jones of the Agentic AI Foundation, a standards body, says engineers must learn to build AI agents, and companies need their own rules for the tools agents connect to.

Bill Doerrfeld, a technology journalist, interviews Angie Jones, a vice-president at the Agentic AI Foundation, which hosts open standards for AI agents. Jones says engineers use AI agents and too few are learning to build systems around them. Doerrfeld reports that use of the Model Context Protocol, a standard for connecting agents to data and tools, is growing faster than companies' rules for it.

• Jones names the skills she finds missing: controlling what an AI agent can and cannot do, observing its actions, and tracing a failure to debug it.
• Claude, Anthropic's AI assistant, has served more than one billion tool calls through the Model Context Protocol, according to David Soria Parra, who helped create the protocol.
• For Jones, downloading a Model Context Protocol server from the internet is not secure practice, and engineers must decide how such servers are governed.
• She advises senior leaders to ask whether a standard is stable, governed by a neutral body and open to their contributions before building on it.

An engineering team can decide which Model Context Protocol servers its AI agents may use before use spreads; the timing is an inference and is not Jones's claim. Jones works for the foundation that hosts the standards she recommends.

Derived by Working Surface from the article; more in the Playbook. Source line: I don’t see enough engineers picking up those skills

10 October 2026: as AI agents start work without being asked, the rules that limit them need to be written down and owned: by a person, a team and the machine itself.