Posts

Showing posts with the label code-review

Copy This Brief: A Coordinator Agent Plus Cross-Vendor Review

Image
A written brief at the center, with review coming in from the sides. In Green Tests, Broken App , I changed how fixes get built: one AI coordinator plans the work, builders write the code, and LLMs from other vendors review every piece. In The Bug Was in the Instruction, Not the Code , those reviewers caught the coordinator's own instructions being wrong. Those posts are the stories. This one is the part you can copy: the brief, the builder instruction, and the review checklist. The templates are a starting point. I have not measured them, and you do not need my exact setup to use them. Who Does What Think of a home remodel. The coordinator is the general contractor. It is one AI session that plans the work, writes an instruction for each builder, checks review findings, and puts the pieces together. It writes almost no code. A lane is one room: one slice of the change with its own files, so two crews are never tearing out the same wall. A builder is the crew in that r...

The Bug Was in the Instruction, Not the Code

Image
Three builders, one instruction, and a crack in the instruction. In one release, the reviewers caught the coordinator's own instructions being wrong three times. Not the builders' code. The instructions. I use AI agents to fix my assistant app. In Green Tests, Broken App , I ended with a rule: one coordinator plans and checks the work, builders from three vendors write the fixes, and every piece is checked by the two vendors that did not build it. Between October 27 and November 2, I ran that rule through eight more releases. The builders' mistakes were caught, as before. What I did not plan for was that the coordinator's mistakes got caught too. I count them as mine, because I run the coordinator. The Setup The coordinator is one Claude Opus 5.5 session. It plans each release, writes a short instruction for each builder, checks every review finding against the real code before passing it on, puts the pieces together, and ships. The builders are Claude, Codex runn...

Green Tests, Broken App: One Coordinator and Three AI Agents

Image
A smoke detector with a green light is not proof that it works. The light only says the battery is in. To know, you press the button and listen. This post is about the week I stopped trusting the green light on my own app, pressed the button, and changed how the fixes get built. In my last post, Five Reviews in One Release , I ended with one question to ask about every fix: what else does this touch? I also said I would stop letting a green checkmark tell me I am done. This post is the bigger version of that lesson. It was not one fix that fooled me. It was the whole app. The Setup My personal assistant app is the one I wrote about in Building a Personal Assistant: What I Got Wrong . Every change to it comes with tests: small programs that check the app does what it should. By October 22, 905 issues were closed, 422 of them since August 1. Nearly all were closed because about 4,700 tests passed. That day I said it plainly: basic things did not work, and I had lost trust in the a...

Five Reviews in One Release, and Two Fixes That Created the Next Bug

Image
You fix a squeaky door, and now it hits the light switch when it opens. Anyone who has done a home repair knows this feeling. The fix was right. It just touched something you were not looking at. This post is about how often that happens when AI writes my code, and what my current way of working catches and misses. The Setup In Seven Releases With Grok , Grok wrote most of the code and Claude reviewed it. I ended that post saying I would spend my Grok usage on reviewing more than on writing code. So for this round I turned the roles around. Claude Opus 5.5 wrote the code. Two models from other companies reviewed it: GPT-6 Astra first, then Grok 4.7. This is the idea from Executor Model, Reviewer Model : the model that checks the work should not come from the same company as the model that did it. The work was a feature for my personal assistant app. In a long chat, you can now pin a message, or pick up a question and its answer and move them into a new conversation of their own. ...

Seven Releases With Grok: Fast, Honest, and What the Diff Did Not Show

Image
In my last post I cancelled Ollama. The money running out early was enough on its own, but the part I wrote about was how many review rounds the work needed before I could merge it. That gives me a simple test for any model: how many times does the work go around review, and does each fix pass make things better or add new problems? This post runs the same test on Grok. Starting on September 21, it implemented four releases of my Zone 2 Trainer app and two releases of my personal assistant app. On a seventh release I swapped the roles, and Grok reviewed code that Claude wrote. Grok did well on that test. That is the short version. The longer version is about where the bugs that did get through were hiding: not in what the changed lines did, but in what they assumed. The Setup The models are grok-4.6 and grok-4.7 from xAI, on a SuperGrok subscription. grok-4.7 came out right after the first release, so it did the rest. On the other side was Claude. It wrote the issue lists and...

A Convincing Review of the Wrong Path

Image
This article follows Executor Model, Reviewer Model: Why Cross-Vendor Code Review Actually Works , where I wrote about using a second model to validate review findings. This time, the missing check was not another opinion. It was running the code. I am the only user of my personal assistant app, so every issue in its tracker is one I filed after watching something fail on my phone. One was the fasting timer. I told it in chat that I had started a fast at seven in the evening, and nothing started. I handed that issue and eight others to Claude. It read the source, quoted the right files, and explained why the timer could not fail the way I described. I closed all nine issues that afternoon. A few days later I asked Claude to check again, but this time it had to run the reported inputs. Four verdicts flipped. The second pass took six and a half minutes instead of two and a half. Those extra four minutes recovered four real bugs and found two more I had never noticed to file. T...