From 22da10941c0a8fc0bcdd35072286dba7ebeef233 Mon Sep 17 00:00:00 2001 From: Charles Ponti Date: Tue, 20 Oct 2026 03:43:11 +0100 Subject: [PATCH] fix(skills): chase-review allows Anthropic models in Thames's picker --- skills/chase-review/references/data-and-trust.md | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/skills/chase-review/references/data-and-trust.md b/skills/chase-review/references/data-and-trust.md index 8475a39..adae9fe 100644 --- a/skills/chase-review/references/data-and-trust.md +++ b/skills/chase-review/references/data-and-trust.md @@ -37,9 +37,10 @@ - Validate at the edge with a schema; return typed errors. No trusting the client. - Reply costs, rate limits and paid calls are visible in code and docs; say what a live test costs. - Models and vendors: keep the model list in one place, check live model availability and - deprecation dates before choosing, and prefer cheap fast defaults. Never an Anthropic model in - Thames's picker (word, 2026-10-13). A saved choice that is no longer listed falls back to the - default. + deprecation dates before choosing, and prefer cheap fast defaults. Any vendor is allowed in + Thames's picker, Anthropic included (word, 2026-10-20; it was banned 2026-10-13). Benchmark + candidates with the app's real request before picking, and confirm each passes the live test. A + saved choice that is no longer listed falls back to the default. ## Dev services - You may start local dev services (Metro, `pnpm dev`, API, workers, databases, Docker) to verify a