Nobody had to ask for the logs
keryx has no users but me, and it still got seven bug reports in two days, all filed by one of my own agent sessions. Checking one meant reading exactly what the reporter had typed, which you don't normally get to do.

keryx has no users but me, and it still got seven bug reports in two days, all filed by one of my own agent sessions. Checking one meant reading exactly what the reporter had typed, which you don't normally get to do.

Two weeks tuning a retrieval system, every day productive and the numbers still not good enough... so I finally went and read what other people had already built.

Twenty agent sessions, one GitLab account on the Free plan and a rate limit that kept tripping. The swarm was the obvious suspect, right up until I traced it.

The tool had comments on under ten per cent of its lines and ninety pages of documentation built into the binary, and I'd argue that was one decision made for two very different readers.

Agents made the doing cheap. What fourteen sessions cost is walking into each one cold, and the one head doing the walking.

A model handed a bug will hand you back a plausible fix and a green test suite in about a minute. That tells you almost nothing.

I read the question my support bot had fluffed, gave the correct answer off the top of my head, and was wrong. The docs had fooled the person who wrote them.

Eleven variants of one shell script, on hundreds of laptops, launching pods into production. The language was never the problem, and it wasn't the fix either.

Two configs for one tool, the user's and the project's, and no rule anywhere about which one wins. Twenty minutes to make one a layer of the other and keep a straight answer about where every value came from.

go/mcp puts a Go CLI, HTTP service or gRPC service on the Model Context Protocol with three discovery tools instead of one per command, so a model carries a lantern rather than the whole catalogue.
