Notes on running parallel development with Claude Code by splitting it into separate personas — the configuration actually in use, and the failures behind it.
I had an MCP server that occasionally hung, so I gave it a short leash:
I put autoMode in .claude/settings.json, committed it, and it had no effect.
I wanted to stop auto-approving a command, so I deleted its entry from permissions.allow in my ~/.claude/settings.json.
I had a rule meant to keep an agent inside one directory:
I run Claude Code as five separate subagents with different permissions — architect, coder, reviewer, conflict resolver, UI designer. One of them sto…
I have a job that publishes one article a day. Over four days I rebuilt its scheduling three times, and each rebuild fixed a real failure that the pr…
I maintain a linter for a configuration format. Its stated principle is on the first screen of the README:
I published three CLIs of my own to npm. All three turned out to be broken, found in the last few minutes before publishing.
I shipped a bug in a script I sell. I had also reported that script as "verified across four scenarios."
You write a hook. The JSON is valid. The keys are spelled right. No error appears.
I wanted to add links to nine already-published articles. Just a bulk update through my own publishing tool.
There are two storefronts, one domestic and one international, and the international one is not open yet.
I built a checker for a configuration directory. The time went not into adding rules, but into deciding what not to add.
I wrote a static site generator and added canonical and hreflang to 28 article pages. Fourteen in Japanese, fourteen in English. Until then, search e…
I wrote a tool that reads GitHub Issues and reports the ones whose declared file scopes overlap.
I built a pipeline that publishes articles to Zenn by git push, using its GitHub integration. I pushed twelve at once. Two went live. The other ten d…
I publish technical articles to Zenn, to dev.to, and to my own site. Same content. Done naively, that leaves Google looking at the same article three…
Writing "Write clean code" in CLAUDE.md changes nothing.
Has this ever happened to you when trying to run multiple tasks in parallel?
When using cloud-based LLMs, token costs become a concern. It is natural to think, "Can I offload light tasks to a local LLM to save money?"
When tracking progress by adding labels like status:in-progress or status:review to Issues, you will inevitably forget to apply them.
When you create a small CLI tool, it is common to consider publishing it on npm for others to use. However, there is a distribution path that require…
This article is for those who encountered issues when running multiple tasks simultaneously with Claude Code.
When you give a large request to Claude Code, doesn't this happen?
When you ask generative AI to create a screen, these patterns appear every time.
I have a question for those who are using AI to write code.