From 34bbfde2f41ebed0faef36dd3420ee652b4c9d74 Mon Sep 17 00:00:00 2001 From: can1357 Date: Mon, 27 Jul 2026 22:44:48 +0200 Subject: [PATCH] feat(coding-agent/modes): added tip for cycling reasoning effort levels - Add a tip informing users about pressing shift+tab to cycle through reasoning effort levels. --- packages/coding-agent/src/modes/components/tips.txt | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/packages/coding-agent/src/modes/components/tips.txt b/packages/coding-agent/src/modes/components/tips.txt index 7e8c60848..d7de30404 100644 --- a/packages/coding-agent/src/modes/components/tips.txt +++ b/packages/coding-agent/src/modes/components/tips.txt @@ -22,4 +22,5 @@ Press ← ← to drill into a running or finished agent and inspect its tool cal Hit a Codex rate limit? `/usage reset` spends a saved reset credit to immediately restore your quota No native tool_calling? Inference provider botches parsing them? `PI_DIALECT=glm|kimi|anthropic…` rolls it locally for them! Turn on `/advisor` to attach a second model that reviews every turn and quietly injects advice -Try starting your prompt with a ->, and writing a list (1. Do X, 2. Do Y) \ No newline at end of file +Try starting your prompt with a ->, and writing a list (1. Do X, 2. Do Y) +Press shift+tab to cycle through reasoning effort levels \ No newline at end of file