From d18b064e43d1460da7ff2e3d9d6416e7fc88b749 Mon Sep 17 00:00:00 2001 From: Lilleman auf Larv Date: Wed, 21 Oct 2026 21:41:31 +0200 Subject: [PATCH] =?UTF-8?q?Answer=20architecture=20round=202:=200=20?= =?UTF-8?q?=F0=9F=94=B4,=200=20=F0=9F=9F=A1;=20cite=20the=20synced=20effor?= =?UTF-8?q?t=20without=20the=20model?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- docs/decisions.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/decisions.md b/docs/decisions.md index e2ee8ecf..d12f75df 100644 --- a/docs/decisions.md +++ b/docs/decisions.md @@ -164,7 +164,7 @@ Three places carry the rule. `rules/todo-items.md` states it and loads with `tod ## settings.json: each machine owns its model, and git's clean filter `bin/settings_filter.py` keeps it, and Claude Code's key order, out of every commit -2026-10-21, lilleman. Serves goal 2: Claude Code rewrites `"model"` in `settings.json` on every `/model` and plugin install, so a tracked model let one machine's pick override every other's. lilleman chose a model per machine over a synced one in `env.ANTHROPIC_MODEL`, which `/model` could not outlast past one session: it is simpler, and loop-claude picks each project's model in `~/.claude/loop-claude-models.json` anyway. The filter also sorts the keys, since each write puts them back in Claude Code's order. Git still lists `settings.json` after a pick that changes the file's size, and `git pull` refuses to update it, until `git add settings.json`, which stages no change. Until then, Update's `git checkout -f` rewrites the file and drops the model, so Update stages it first. A move or pull that changes `settings.json` drops the machine's model either way. The effort keys, `effortLevel` and `modelSettings`, stay synced: the routing decision on a chunk's worker rung relies on "`settings.json` runs Opus 5.5 at xhigh". Valid while Claude Code writes the model it picks into `settings.json`. +2026-10-21, lilleman. Serves goal 2: Claude Code rewrites `"model"` in `settings.json` on every `/model` and plugin install, so a tracked model let one machine's pick override every other's. lilleman chose a model per machine over a synced one in `env.ANTHROPIC_MODEL`, which `/model` could not outlast past one session: it is simpler, and loop-claude picks each project's model in `~/.claude/loop-claude-models.json` anyway. The filter also sorts the keys, since each write puts them back in Claude Code's order. Git still lists `settings.json` after a pick that changes the file's size, and `git pull` refuses to update it, until `git add settings.json`, which stages no change. Until then, Update's `git checkout -f` rewrites the file and drops the model, so Update stages it first. A move or pull that changes `settings.json` drops the machine's model either way. The effort keys, `effortLevel` and `modelSettings`, stay synced: the routing decision on a chunk's worker rung relies on `settings.json` running every interactive session at xhigh effort. Valid while Claude Code writes the model it picks into `settings.json`. ## the repo may name bender, bender-infra and ai-sandbox