Cloudflare Blog

Cloudflare17d ago

Orchestrating AI Code Review at scale

Cloudflare engineers developed a scalable AI code review system that integrates multiple AI agents to review code quality, security, and compliance. The system uses a plugin architecture to support various version control systems, AI providers, and internal standards. It has been successfully run across tens of thousands of merge requests, approving clean code and flagging real bugs with high accuracy. The system consists of a coordinator process that spawns OpenCode sessions for sub-reviewers and provides configuration through a controlled interface. Each sub-reviewer runs in its own OpenCode session, using its own agent prompt and model. The system isolates VCS-specific coupling and uses a composable plugin architecture to support multiple plugins, including GitLab, AI Gateway, codex, and tracing and observability. The implementation allows for high scalability, flexibility, and maintainability, enabling Cloudflare engineers to review code more efficiently and effectively.

NetworkingSecurity

1 min

Latest Articles

Building for the future

How Cloudflare responded to the “Copy Fail” Linux vulnerability

When DNSSEC goes wrong: how we responded to the .de TLD outage

Code Orange: Fail Small is complete. The result is a stronger Cloudflare network

Introducing Dynamic Workflows: durable execution that follows the tenant

Post-quantum encryption for Cloudflare IPsec is generally available

Agents can now create Cloudflare accounts, buy domains, and deploy

Shutdowns, power outages, and conflict: a review of Q1 2026 Internet disruptions

Making Rust Workers reliable: panic and abort recovery in wasm‑bindgen

Moving past bots vs. humans

Building the agentic cloud: everything we launched during Agents Week 2026

The AI engineering stack we built internally — on the platform we ship

Orchestrating AI Code Review at scale

Introducing the Agent Readiness score. Is your site agent-ready?

Shared Dictionaries: compression that keeps up with the agentic web

Introducing Flagship: feature flags built for the age of AI

Redirects for AI Training enforces canonical content

Agents Week: network performance update

Unweight: how we compressed an LLM 22% without sacrificing quality

Agents that remember: introducing Agent Memory

Cloudflare Blog

Latest Articles

Building for the future

How Cloudflare responded to the “Copy Fail” Linux vulnerability

When DNSSEC goes wrong: how we responded to the .de TLD outage

Code Orange: Fail Small is complete. The result is a stronger Cloudflare network

Introducing Dynamic Workflows: durable execution that follows the tenant

Post-quantum encryption for Cloudflare IPsec is generally available

Agents can now create Cloudflare accounts, buy domains, and deploy

Shutdowns, power outages, and conflict: a review of Q1 2026 Internet disruptions

Making Rust Workers reliable: panic and abort recovery in wasm‑bindgen

Moving past bots vs. humans

Building the agentic cloud: everything we launched during Agents Week 2026

The AI engineering stack we built internally — on the platform we ship

Orchestrating AI Code Review at scale

Introducing the Agent Readiness score. Is your site agent-ready?

Shared Dictionaries: compression that keeps up with the agentic web

Introducing Flagship: feature flags built for the age of AI

Redirects for AI Training enforces canonical content

Agents Week: network performance update

Unweight: how we compressed an LLM 22% without sacrificing quality

Agents that remember: introducing Agent Memory