CI

Evidence (17)

• 6The New Stack·8/7/2026regulation
  The AI kill switch problem is knowing what to shut down

  The article argues that public talk of an "AI kill switch" simplifies a much harder operational problem: identifying exactly which systems, services, and dependencies must be stopped when an AI capability poses unacceptable risk. It notes that as AI moves into production environments, shutdown authority is only useful if organizations can trace and constrain complex, interconnected infrastructure quickly and reliably.

• 6InfoQ·8/7/2026product_launch
  Spotify Details AI Agent for Fleet-Wide Codebase Migrations

  Spotify engineers Jo Kelly-Fenton and Aleksandar Mitic describe Honk, an AI coding agent built to manage large-scale codebase migrations across thousands of repositories. The talk focuses on separating CI verification runtime from the agent, handling pull request bottlenecks, and enforcing standardization across the engineering fleet.

• 7The New Stack·8/6/2026framework_update
  Anthropic pushes per-agent git worktrees, exposing infra bottlenecks

  Anthropic’s guidance to use one git worktree per coding agent makes parallel agent workflows the default, but it also exposes limits in CI, staging, databases, and other shared infrastructure. The article argues that branching now needs to extend beyond code into the runtime stack so each change can be tested end to end.

• 4Hacker News·8/3/2026research
  AI’s Limited Productivity Gains for Developers

  The article argues that AI has improved engineering productivity, but less than many leaders expect because most developer time is spent on planning, debugging, reviews, and coordination rather than writing code. It says junior developers may benefit more than senior engineers, since AI helps most with coding tasks and less with broader software design and decision-making work.

• 5The New Stack·7/31/2026framework_update
  Thoughtworks Engineer Says AI Software Needs Human Oversight On the Loop

  Thoughtworks distinguished engineer Kief Morris argues that teams using AI to build software need humans “on the loop” to define what counts as good code and good systems. He says production-ready CI/CD, tests, and guardrails must be built into agentic workflows now to keep software safe, secure, and maintainable. The article frames continuous delivery practices as the way to control AI agents rather than relying on post-hoc fixes.

• 6The New Stack·7/29/2026research
  AI strategy shifts from coding speed to delivery throughput

  The article argues that many companies invest in AI to make developers code faster, but coding is only a small part of software delivery. It says the biggest gains come from applying AI across the full software development lifecycle, especially in workflows like testing, reviews, incident response, and DevSecOps. The piece cites large regulated enterprises as examples of where AI can deliver measurable value when paired with mature CI/CD and automation.

• 7The New Stack·7/29/2026research
  Shipping Code Without Human Verification

  The article argues that agentic coding is making traditional human code review a bottleneck, especially for verification tasks. It proposes shifting teams toward automated verification systems and a “human on the loop” model, where people maintain the review infrastructure instead of checking every change themselves.

• 5The New Stack·7/25/2026framework_update
  Build AI Context Systems Instead of Fixing Code

  The article argues that software teams should stop trying to correct AI-generated code line by line and instead improve the surrounding system, context, and workflows. It frames this as a broader shift from deterministic software development toward context engineering, where organizations package and scale the information agents need to work effectively.

• 6The New Stack·7/24/2026research
  Test data delays are becoming the main bottleneck for AI-era delivery

  The article argues that AI has accelerated code and test generation, but many organizations still wait too long for production test data, creating a new delivery bottleneck at the validation stage. It cites survey findings that most enterprises rely on manual provisioning steps and need better automation, masking, policy enforcement, and self-service access to compliant test data.

• 6The New Stack·7/22/2026funding
  Anthropic acquires Mendral team to bolster Claude’s software engineering tools

  Anthropic is bringing Mendral’s team in-house in an acqui-hire focused on improving Claude’s software engineering capabilities. Mendral will shut down its hosted product and help customers migrate, while its founders and engineers join Anthropic to work on CI/CD and other developer automation tasks.

• 6The New Stack·7/16/2026product_launch
  GoDaddy opens domain registration to developer and AI workflows

  GoDaddy launched a Developer Platform that lets teams search, buy, configure, and manage domains from within development tools instead of using the registrar web dashboard. The release includes APIs, a CLI, OAuth-based access controls, and purchase safeguards such as quote-then-execute flows, idempotency keys, and human approval for automated or AI-driven actions.

• 3The New Stack·7/14/2026product_launch
  Tailscale webinar explores access control for AI agents and humans

  The New Stack article promotes a webinar about how human-focused access tools struggle when organizations add AI agents, CI/CD jobs, and other non-human identities. It highlights the need for a unified access architecture that combines network access control, PAM, and policy-based revocation and auditing under one model.

• 6The New Stack·7/11/2026research
  AI code reviews expose a new software bottleneck

  The article argues that merge quality should be treated as an explicit contract, especially as coding agents increase pull request volume. It says traditional pipelines usually verify only the first three layers of confidence, while real system behavior testing remains the costly missing layer that often catches microservice bugs.

• 3Hacker News·7/7/2026product_launch
  Service to Reduce AI-Generated Codebases

  A Hacker News post advertises a paid service that audits and shrinks large AI-generated codebases, promising a fixed reduction target, a one-week engagement, and post-work guardrails. The team says it uses Claude Code carefully and focuses on preserving functionality while removing duplicated or unmaintainable code.

• 8The New Stack·7/1/2026security
  Cordyceps flaw shows CI/CD is part of the attack surface

  Research from Novee Security describes the “Cordyceps” CI/CD weakness, which could let unauthenticated GitHub accounts hijack trusted workflows and compromise open-source supply chains. The report says it found hundreds of potentially exploitable repositories, reinforcing that CI/CD pipelines should be treated as security-critical production systems rather than mere configuration.

• 8The New Stack·7/1/2026security
  Codecov shows how CI pipelines can become the attack surface

  The article argues that the Codecov breach was not an isolated incident but an example of a broader pattern in which attackers target software build and deployment pipelines. It cites later supply-chain compromises such as XZ Utils and polyfill.io to show that CI/CD systems and other trusted delivery mechanisms can be used to exfiltrate secrets or deliver malware at scale.

• 6The New Stack·6/23/2026research
  Kubernetes teams trust automation for delivery but hesitate on CPU and memory tuning

  A survey of 321 enterprise Kubernetes practitioners found strong trust in automated deployment and rollback, but much lower willingness to let systems auto-apply CPU and memory resource changes. The gap matters more as AI inference workloads on Kubernetes raise the cost of overprovisioning and make manual rightsizing harder to sustain at scale.