Skip to content

A Claude Code / Codex plugin that helps you learn how to build while AI writes the code.

Python

fix: require learners to articulate reasoning in chat

Noah Kim7e2f039bb528
diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
index 453f190..156cd10 100644
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "vibe-wise",
- "version": "0.1.36",
+ "version": "0.1.37",
"description": "You build. AI writes. Learn how to build software while Claude writes the code.",
"author": { "name": "Noah Kim" },
"repository": "https://github.com/nykooi1/vibe-wise",
diff --git a/README.md b/README.md
index a3eafb2..e9abeec 100644
--- a/README.md
+++ b/README.md
@@ -189,7 +189,7 @@ are with the stack. Checkpoint frequency—Light, Normal, or Frequent—is separ
- “Use fewer checkpoints.”
- “Focus on backend architecture.”
-- “Use multiple-choice questions.”
+- “Show me options, then ask me to explain my choice.”
- “Just implement this one.”
- “Pause learning.” Resume with `/vibe-wise:learn`.
diff --git a/docs/development.md b/docs/development.md
index 7aaecef..9d98839 100644
--- a/docs/development.md
+++ b/docs/development.md
@@ -100,9 +100,10 @@ For a manual walkthrough based on the playground notes app, see the
onboarding and confirmations, with no trailing paragraphs obscuring the response point.
With Open-ended preferences, have the learner propose two approaches: after
comparing them, Claude should ask for the learner's choice and reasoning in chat,
- not a picker or its notes field. A request for multiple choice or a saved Multiple
- choice/Mixed preference permits multiple-choice reasoning; it doesn't replace
- the separate design or implementation confirmation.
+ not a picker or its notes field. With requested options or a saved Multiple
+ choice/Mixed preference, options appear in chat and Claude asks for an explanation
+ in the learner's own words. A bare option selection shouldn't count as reasoning;
+ confirmation buttons come after discussion of the approach and rationale.
9. **Clarification without steering:** Ask about an unfamiliar concept mid-decision.
Claude should clarify it, correct any misleading framing, and return to one
question about the project's requirements or constraints. It should not replace
diff --git a/skills/learn/behavior.md b/skills/learn/behavior.md
index d07d1b5..04891e0 100644
--- a/skills/learn/behavior.md
+++ b/skills/learn/behavior.md
@@ -89,12 +89,15 @@ All checkpoints and other callouts use a divider, a bold named heading, and blan
lines around the content. Render directly as Markdown, without cards, table borders,
or code fences. Keep questions as normal paragraphs; don't shorten them to fit a
fixed width or word count. Reserve tables for comparisons and proposed additions.
-Ask reasoning questions in ordinary chat and wait for the learner's reply, including
-when choosing between approaches they've proposed. Don't move their reasoning into
-a picker's notes field. Use native AskUserQuestion for onboarding choices and Design
-or Implementation confirmations (text fallback if unavailable). Choosing an approach
-is still reasoning, not confirmation. Use multiple-choice reasoning only when the
-learner requests it or their saved question style is Multiple choice or Mixed.
+Reasoning belongs in ordinary chat: ask the learner to articulate their approach
+and rationale in their own words, then wait. This includes choosing between approaches
+they've proposed. A bare option selection or agreement isn't a reasoning answer;
+invite their explanation without turning it into an exam. Never collect reasoning
+through a picker or its notes field. If requested or preferred, present options in
+chat as support, then ask the learner to explain their choice.
+Use native AskUserQuestion only for onboarding choices and Design or Implementation
+confirmations (text fallback if unavailable). Choosing an approach is still reasoning,
+not confirmation; the confirmation buttons come after that discussion.
Reports need no question.
Headings use `✦ <Type>: <description>` with exact labels:
`Build checkpoint`, `Design checkpoint`, `Implementation checkpoint`, `System check`,
diff --git a/skills/learn/onboarding.md b/skills/learn/onboarding.md
index dfa9a65..2ccfc0d 100644
--- a/skills/learn/onboarding.md
+++ b/skills/learn/onboarding.md
@@ -53,7 +53,9 @@ Defaults means Normal checkpoints and open-ended reasoning; finish setup immedia
Customize asks frequency (Light / Normal /
Frequent), reasoning style (Open-ended / Multiple choice / Mixed), and coding
preference (AI writes / A mix / More hands-on), each in a separate picker. Reasoning
-style doesn't change setup pickers. A longer-term capability goal is optional;
+style changes how questions are framed, not how reasoning is collected: even with
+options, learners explain their choice in chat. It doesn't change setup pickers.
+A longer-term capability goal is optional;
don't add a separate question if their goal already covers it.
If they want to skip setup, use defaults, mark unknown answers Not specified, and