Intel / blog

The Agent Development Lifecycle has arrived on Cloudflare

Source: blog.cloudflare.com
Author: Brendan Irvine-Broque
Date: Published Aug 4, 2026

Context

Cloudflare’s proposed Development Lifecycle addresses the work that remains after an agent generates code. Testing, deployment and maintenance require interfaces an agent can call, evidence it can inspect and permissions that constrain its actions. The company’s proposal uses Workflows to coordinate steps, retain progress and retry failures, with agents invoked where judgment is needed.

Its issue-triage example gives this direction a concrete form: automate investigation and produce a preview for the reporter to verify. That illustrates a bounded workflow within the broader ambition to automate software delivery.

Key takeaways · AI-distilled

AI made implementation — historically the slowest SDLC phase — the fastest and cheapest, which now overwhelms every downstream step: maintainers flooded with PRs, production engineers racing the delivery rate.
The paradoxical fix is giving agents MORE of the lifecycle, not less: most teams let agents write code while humans still validate, merge, deploy, and hold the pager — a division no engineering team would accept for a human.
For agents to own the lifecycle, every step must become programmatic (no ClickOps), horizontally scalable (a preview per agent), reproducible, push-based, atomic, and permissioned.
Concrete tooling shipped alongside the thesis: @cloudflare/ci self-healing CI/CD on Workflows, OpenTelemetry traces in local dev, and agent-trace observability.
Proof points from Cloudflare's own use: AI-enforced engineering standards across its repos, and a software factory that drove Astro's GitHub issue count to zero.

Terms in this piece · Glossary

AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.

Key quotes

“AI has made the step that was previously the slowest and most expensive — implementation — the fastest and cheapest.”

“You’d never let an engineer on your team write code, expect someone else to validate it, merge it, deploy it, hold the pager in production, and triage incoming bugs. But that’s what most companies are doing right now with agents.”

“We think it’s time to replace the SDLC with the ADLC — the Agent Development Lifecycle.”

“Software factories face the same challenge that other autonomous systems like self-driving cars do — the challenge of going from working successfully 80% of the time, to some number of nines past 99%.”

“Ask yourself — why haven’t you yet just let your agent auto-approve and merge its own PRs to your production services?”

Read the source blog.cloudflare.com