3 agentic AI questions engineering leaders must answer
Jennifer Riggins · 10 November 2026
Read the originalJournalist Jennifer Riggins asked developer-experience researchers how to measure AI agents' effect on engineering teams; they point to wellbeing, enforced accountability and the cost of checking agents' work.
With no DORA survey of software delivery performance this year, LeadDev asked researchers which measures engineering organisations need now that AI agents do more of the work. Jennifer Riggins, a freelance technology journalist, gathered answers from researchers at AWS, DX, Octopus Deploy and elsewhere. They agree that the volume of code says little and that rules for agents must be enforced, not only written down.
- Nathen Harvey, formerly DORA's lead, expects developer wellbeing to fall as agents add context switching and raise expectations of output. One survey found 41% of developers show signs of being hooked on AI coding.
- Although 87% of organisations claim AI governance policies, fewer than a quarter have put controls in place. Charlotte Fleming of Octopus Deploy argues that the rules must run as gates in the deployment pipeline.
- Laura Tacho of AWS recommends watching rework rate, code review wait time and pull request size, alongside how developers rate the quality of agent-written code.
- A June 2026 McKinsey report found that 60% of agentic AI spending went to checking and correcting output, a cost that deployment frequency does not show.
An engineering team using agents could turn its rules for what agents may do alone into checks in the deployment pipeline, and track rework and review time. The piece gathers expert opinion and reports no single team's results.
Turn the team's rules on what AI agents may change without a person's approval into automated checks in the deployment pipeline that block any change breaking them.
Track rework rate, code review wait time and pull request size as AI agents take on more of the coding.
Derived by Working Surface from the article; more in the Playbook. Source line: A policy document states what you intend to happen. A deployment gate actually makes it happen.
13 November 2026: AI agents earn more freedom when a team limits what they can reach, from data to design components, and adds a check that does not depend on a glance.