From 33c43bbf2bd1df4969e81be850acd77b8b0eadc9 Mon Sep 17 00:00:00 2001 From: Kyle Sexton <153232337+kyle-sexton@users.noreply.github.com> Date: Sun, 1 Nov 2026 21:49:55 -0500 Subject: [PATCH] revert(naming): restore the 650-character name-it-better description The 624-character cut shipped in #12186 on an eval that cannot tell it from "Helps with naming.": on Sonnet 5.5 the generic description scored 0.983 vs 0.967 on routing and fired 24/24 should-trigger probes, the same as the shipped text. Under the negative-control rule the cut was unmeasurable, so the earlier text returns until a suite can measure it. Refs: #6611 Co-Authored-By: Claude Opus 5.5 --- .../revert-6611-name-it-better-description-b4e2c7a1.md | 7 +++++++ plugins/naming/skills/name-it-better/SKILL.md | 2 +- 2 files changed, 8 insertions(+), 1 deletion(-) create mode 100644 .changes/naming/revert-6611-name-it-better-description-b4e2c7a1.md diff --git a/.changes/naming/revert-6611-name-it-better-description-b4e2c7a1.md b/.changes/naming/revert-6611-name-it-better-description-b4e2c7a1.md new file mode 100644 index 00000000000..66fa9c278bb --- /dev/null +++ b/.changes/naming/revert-6611-name-it-better-description-b4e2c7a1.md @@ -0,0 +1,7 @@ +--- +bump: patch +--- + +### Changed + +- The `name-it-better` description is back to its 650-character text. The shorter one shipped on an eval that a negative control showed cannot tell it from a one-line generic description, so the cut was never measured. diff --git a/plugins/naming/skills/name-it-better/SKILL.md b/plugins/naming/skills/name-it-better/SKILL.md index 1873300905b..2f20ffb4b66 100644 --- a/plugins/naming/skills/name-it-better/SKILL.md +++ b/plugins/naming/skills/name-it-better/SKILL.md @@ -1,5 +1,5 @@ --- -description: "Generate and evaluate fresh name candidates for anything: a variable, function, file, module, skill, repo, or domain term. The human picks; never auto-locks a name. Use when the target name is still UNDECIDED: a fresh name ('suggest names', 'what should I call this') or a retry on one just rejected ('name it better', 'that name is wrong', 'help me rename this to something better'). Not for an already-decided rename ('rename X to Y', 'I renamed X'), which routes to the rename-references sweep. Spawns blind fresh-context generators from distinct lenses. Optional 'tournament' arg for high-stakes, hard-to-refactor names." +description: "Generate and evaluate fresh name candidates for anything: a variable, function, file, module, skill, repo, or domain term. Then let the human pick. Use when the target name is still UNDECIDED, whether the ask is a fresh name ('suggest names', 'what should I call this') or a retry on one just rejected ('name it better', 'that name is wrong', 'help me rename this to something better'). Not for an already-decided rename ('rename X to Y', 'I renamed X'). That routes to the rename-references sweep. Spawns blind fresh-context generators from distinct lenses; never auto-locks a name. Optional 'tournament' arg for high-stakes, hard-to-refactor names." argument-hint: "[tournament]" user-invocable: true disable-model-invocation: false