I'm Ben Nevile. I've launched many businesses and software platforms, and I'm based in Montréal.
Projects
Recent notes
-
Little Pan: most review bots never set GitHub's review decision, so auto-merge does not wait for them
Oct 26, a how-to on waiting for CI and review on agent-opened PRs finds that Codex, Claude Code Review, Copilot and Gemini mostly comment or post neutral checks. Auto-merge waits only for required checks, and with none required it merges at once.
-
Samuel Medeiros: four ways a green agent loop lies, and a delivery gate that measures instead of asking
Oct 26, Samuel Medeiros on a run where the builder said done and the gate scored 55 with a token in the diff. His rule is builder, critic, and reviewer in separate contexts, with a fail-closed security gate, and he lists four ways a passing check can hide a failure.
-
Conor Bronsdon: when agents write most of the code, the test suite is the gate and people sign off at the boundary
Drawing on Graphite's Greg Foster and AMD's Anush Elangovan, a Chain of Thought explainer argues that review at agent volume moves from reading every line to rerunnable tests, small stacked diffs, and human sign-off on what enters and leaves the agents' working enclosure.
-
John Crowley (Port): a factory that only runs tasks ships the same quality in month six as in week one
Oct 25, John Crowley at Port on closing the loop in an AI software factory: capture each run, grade a sample on a fixed rubric, find the pattern across a batch, and write a reviewed fix back into the skill the next run reads.
-
Ahmad Dogar: agent reports must carry evidence, and the section he reads first is what it did not check
Oct 25, Ahmad Dogar on making coding agents prove their work: checks ordered cheapest first, a fixed report format with pasted output, a Stop hook that keeps the agent working while typecheck or lint is red, a bundle budget printed on every run, and screenshots opened by someone who did not write the code.
-
Hiroki Kondo: 43 of 80 AI reviews were re-reviews, and a review step written down for agents ran once in about 290 sessions
Oct 25, the third weekly log for karasu, built entirely through Claude Code. 50 PRs merged in a week. Re-reviews on every fix burned the CodeRabbit quota, so he moved a local review ahead of the web one. When agents skipped that written step almost every time, he turned it into hooks that refuse to mark a PR ready.
-
Ted Factory: a PR agent that merges every ten minutes, and why its owner chose thicker review over putting a human back
Since July, a self-built agent has reviewed, merged, resolved conflicts, and deployed across one company's repositories on a 10-minute loop. Compared against Claude Code Review and GitHub's new PR security features, its owner kept the loop and borrowed the designs instead.
-
Wakim's night shift: agent jobs built, reviewed and merged by 1:40 a.m., after a four-day outage nobody noticed
Oct 25, Charbel Wakim on running headless Claude Code jobs overnight with a reviewer agent and a merge step. The scheduler kept failing quietly: usage limits, a sleeping laptop, an exit message that said a finished job had died. What made the night shift trustworthy was a row each job updates after every step.
-
Zontak's one-Mac factory: a ticket became a reviewed PR overnight, and the limit he names is his own attention
Oct 25, Denis Zontak writes up one sleepless night building a small software factory on a Mac and a $200 subscription: plan, plan-check, sandboxed build, 1,100 tests, diff review, then his sign-off. The limits he lists are usage caps, review attention, and what the checks can test.
-
Hayden Bleasel: code review was always two jobs, and agents now do the bug-finding one better
An OpenAI engineer splits review into objective checks (does it work) and subjective ones (is it the right code), argues agents should own the first, and says consequences, not the split, decide how much to trust them alone.
-
Redman: a nightly agent pipeline's final check passed for months because it was testing the wrong branch
A 20-agent overnight pipeline reported green every morning while its last verification job re-tested a tree that had already passed. Binding each verdict to file hashes and re-checking them at merge then refused all 23 open agent PRs.
-
Simple Memo: a pull request shipped at 08:06, and the pipeline's daily report said nothing did
A scheduled content pipeline merged only the exact commit it had verified. The run record landed on the same branch after the merge, so it never reached main, and the human-facing report counted a zero day. 17 of 41 after-the-fact records had stranded the same way.