claude-plugin-viban 1.0.24 → 1.0.26

package/README.md

@@ -50,89 +50,44 @@ This separation keeps your workflow clean and prevents context switching.
 ## Requirements
 
 - zsh
-- [gum](https://github.com/charmbracelet/gum) - `brew install gum`
-- [jq](https://jqlang.github.io/jq/) - `brew install jq`
+- [gum](https://github.com/charmbracelet/gum)
+- [jq](https://jqlang.github.io/jq/)
 
-## Installation
-
-### One-liner (Recommended)
-
-Installs all dependencies (zsh, gum, jq) and viban automatically:
-
-```bash
-curl -fsSL https://raw.githubusercontent.com/happy-nut/claude-plugin-viban/main/install.sh | bash
-```
-
-### Via npm
+> **Tip:** If using Claude Code, run `/viban:setup` to install all dependencies automatically.
 
-If you already have zsh, gum, and jq installed:
-
-```bash
-npm install -g claude-plugin-viban
-```
+## Installation
 
-After installation, `viban` command is automatically available in your terminal.
+### For Claude Code Users
 
-**Verify installation:**
 ```bash
-viban help
+/plugin marketplace add https://github.com/happy-nut/claude-plugin-viban
+/plugin install viban
+/viban:setup   # Installs all dependencies automatically
 ```
 
-### Shell Setup (if viban command not found)
-
-If `viban` is not found after npm install, add npm global bin to your PATH:
+### For Terminal Users
 
-**For zsh (macOS default):**
 ```bash
-# Add to ~/.zshrc
-export PATH="$PATH:$(npm config get prefix)/bin"
-
-# Reload
-source ~/.zshrc
-```
-
-**For bash:**
-```bash
-# Add to ~/.bashrc
-export PATH="$PATH:$(npm config get prefix)/bin"
-
-# Reload
-source ~/.bashrc
+curl -fsSL https://raw.githubusercontent.com/happy-nut/claude-plugin-viban/main/install.sh | bash
 ```
 
-### Manual Installation
-
+Or via npm (requires zsh, gum, jq pre-installed):
 ```bash
-git clone https://github.com/happy-nut/claude-plugin-viban.git
-cd claude-plugin-viban
-chmod +x bin/viban scripts/check-deps.sh
-
-# Option 1: Symlink to /usr/local/bin
-ln -s "$(pwd)/bin/viban" /usr/local/bin/viban
-
-# Option 2: Add to PATH
-echo 'export PATH="$PATH:'$(pwd)'/bin"' >> ~/.zshrc
-source ~/.zshrc
+npm install -g claude-plugin-viban
 ```
 
-### Claude Code Plugin
-
-To use viban commands in Claude Code:
+<details>
+<summary>Troubleshooting: viban command not found</summary>
 
+Add npm global bin to your PATH:
 ```bash
-# Add as plugin marketplace
-/plugin marketplace add https://github.com/happy-nut/claude-plugin-viban
-
-# Install the plugin
-/plugin install viban
-
-# Install dependencies (first time only)
-/viban:setup
+# For zsh
+echo 'export PATH="$PATH:$(npm config get prefix)/bin"' >> ~/.zshrc && source ~/.zshrc
 
-# Now you can use:
-/viban:assign
-/viban:task
+# For bash
+echo 'export PATH="$PATH:$(npm config get prefix)/bin"' >> ~/.bashrc && source ~/.bashrc
 ```
+</details>
 
 ## Usage
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "claude-plugin-viban",
-  "version": "1.0.24",
+  "version": "1.0.26",
   "description": "Terminal Kanban TUI for AI-human collaborative issue tracking",
   "main": "bin/viban",
   "bin": {

package/scripts/generate-release-notes.sh

@@ -0,0 +1,86 @@
+#!/bin/bash
+# Generate release notes from conventional commits
+# Usage: ./generate-release-notes.sh [previous_tag] [current_tag]
+
+set -e
+
+PREVIOUS_TAG="${1:-$(git describe --tags --abbrev=0 HEAD^ 2>/dev/null || echo "")}"
+CURRENT_TAG="${2:-$(git describe --tags --abbrev=0 HEAD 2>/dev/null || echo "HEAD")}"
+
+# If no previous tag, get all commits
+if [[ -z "$PREVIOUS_TAG" ]]; then
+  COMMIT_RANGE="$CURRENT_TAG"
+else
+  COMMIT_RANGE="${PREVIOUS_TAG}..${CURRENT_TAG}"
+fi
+
+# Arrays for categorized commits
+declare -a FEATURES=()
+declare -a FIXES=()
+declare -a DOCS=()
+declare -a CHORES=()
+declare -a OTHERS=()
+
+# Parse commits
+while IFS= read -r line; do
+  [[ -z "$line" ]] && continue
+
+  # Extract commit message (remove hash prefix if present)
+  msg="$line"
+
+  # Categorize by conventional commit prefix
+  if [[ "$msg" =~ ^feat(\(.+\))?:\ (.+)$ ]]; then
+    FEATURES+=("${BASH_REMATCH[2]}")
+  elif [[ "$msg" =~ ^fix(\(.+\))?:\ (.+)$ ]]; then
+    FIXES+=("${BASH_REMATCH[2]}")
+  elif [[ "$msg" =~ ^docs(\(.+\))?:\ (.+)$ ]]; then
+    DOCS+=("${BASH_REMATCH[2]}")
+  elif [[ "$msg" =~ ^chore(\(.+\))?:\ (.+)$ ]]; then
+    CHORES+=("${BASH_REMATCH[2]}")
+  elif [[ "$msg" =~ ^(ci|test|refactor|perf|style|build)(\(.+\))?:\ (.+)$ ]]; then
+    CHORES+=("${BASH_REMATCH[3]}")
+  else
+    # Skip release commits and merge commits
+    if [[ ! "$msg" =~ ^(Merge|release|chore:\ release) ]]; then
+      OTHERS+=("$msg")
+    fi
+  fi
+done < <(git log --pretty=format:"%s" "$COMMIT_RANGE" 2>/dev/null)
+
+# Generate release notes
+generate_section() {
+  local title="$1"
+  shift
+  local items=("$@")
+
+  if [[ ${#items[@]} -gt 0 ]]; then
+    echo "## $title"
+    echo ""
+    for item in "${items[@]}"; do
+      # Capitalize first letter
+      item="$(echo "${item:0:1}" | tr '[:lower:]' '[:upper:]')${item:1}"
+      echo "- $item"
+    done
+    echo ""
+  fi
+}
+
+# Output release notes
+if [[ -n "$PREVIOUS_TAG" ]]; then
+  echo "# What's Changed"
+  echo ""
+fi
+
+generate_section "Features" "${FEATURES[@]}"
+generate_section "Bug Fixes" "${FIXES[@]}"
+generate_section "Documentation" "${DOCS[@]}"
+generate_section "Maintenance" "${CHORES[@]}"
+generate_section "Other Changes" "${OTHERS[@]}"
+
+# Add full changelog link
+if [[ -n "$PREVIOUS_TAG" ]]; then
+  REPO_URL=$(git remote get-url origin 2>/dev/null | sed 's/\.git$//' | sed 's/git@github.com:/https:\/\/github.com\//')
+  if [[ -n "$REPO_URL" ]]; then
+    echo "**Full Changelog**: ${REPO_URL}/compare/${PREVIOUS_TAG}...${CURRENT_TAG}"
+  fi
+fi

package/skills/task/SKILL.md

@@ -1,6 +1,7 @@
 ---
 name: task
 description: "Analyze problem and register as viban issue with evidence"
+enter_plan_mode: true
 ---
 
 # /task - Problem Analysis and Issue Registration
@@ -10,22 +11,67 @@ Analyze problem situation and register as viban issue with file locations and ev
 > **Core Principle**: Focus on **symptoms and problem definition**, not solutions.
 > Solutions are decided by the assignee after understanding full context.
 
-## Input Verification
+## Interview Phase (Required)
 
-**User Input**: `$ARGUMENTS`
+**Always start with an interview** to gather complete context before exploring code.
 
-If input is empty or unclear:
-1. Use AskUserQuestion to ask about the problem
-2. Proceed after receiving response
+### Interview Questions
+
+Use AskUserQuestion to ask these questions (can combine related ones):
+
+**1. Problem Identification**
+```
+What problem are you experiencing?
+
+- Error message or unexpected behavior
+- What feature/page is affected
+- When did it start happening
+```
+
+**2. Reproduction Context**
+```
+How can this be reproduced?
+
+- Step-by-step actions
+- Environment (local/staging/production)
+- Frequency (always/sometimes/once)
+```
+
+**3. Expected vs Actual**
+```
+What should happen vs what actually happens?
+
+- Expected behavior
+- Actual result
+- Any error messages or logs
+```
+
+**4. Additional Context** (optional, ask if needed)
+```
+Any additional context?
+
+- Recent changes that might be related
+- Workarounds tried
+- Related issues or features
+```
+
+### Interview Strategy
+
+- Ask 1-2 questions at a time using AskUserQuestion
+- Provide options when possible for faster responses
+- Stop interviewing when you have enough to:
+  1. Search for relevant code
+  2. Determine priority and type
+  3. Write clear symptoms
 
 ## Execution Steps
 
-### Step 1: Problem Identification
+### Step 1: Problem Identification (from interview)
 
-Analyze the problem described by user:
-1. **Identify symptoms**: Clearly define what the problem is
-2. **Extract keywords**: Error messages, feature names, module names, etc.
-3. **Determine priority**:
+From the interview responses, extract:
+1. **Symptoms**: Clearly define what the problem is
+2. **Keywords**: Error messages, feature names, module names
+3. **Priority**:
    | Condition | Priority |
    |-----------|----------|
    | System down, data loss | P0 |
@@ -135,47 +181,43 @@ Next steps:
 - `viban start {id}` to start working
 ```
 
-## When Input is Missing
+## Example Interview Flow
 
-Use AskUserQuestion with these prompts:
+**User runs**: `/viban:task`
 
+**Agent**: (AskUserQuestion)
 ```
-What problem should be registered as an issue?
-
-Please include:
-1. What is the problem? (error message, unexpected behavior, etc.)
-2. Where does it occur? (page, API, feature, etc.)
-3. How to reproduce? (step-by-step)
+What problem are you experiencing?
+
+Options:
+- Something is broken/erroring
+- Feature not working as expected
+- Performance issue
+- Need new feature/improvement
+- Other
 ```
 
-## Example
-
-**Input**: "Charts not showing on backtest results page"
+**User selects**: "Something is broken/erroring"
 
-**Analysis Process**:
-1. Keywords: backtest, results, chart
-2. Code exploration:
-   ```bash
-   grep -r "chart" . --include="*.tsx" --include="*.ts"
-   ```
-3. Related file found: `src/pages/backtest/results.tsx`
-4. Check chart rendering logic
-
-**Registration Command**:
-```bash
-viban add "Backtest results chart not displayed" "$'## Symptoms\n...'" P1 bug
+**Agent**: (AskUserQuestion)
 ```
+Please describe the error:
 
-**Registered Issue**:
-```
-Title: Backtest results chart not displayed
-Priority: P1
-Type: bug
-Location: src/pages/backtest/results.tsx
+1. What action triggers it?
+2. What error message do you see?
+3. Which page/feature is affected?
 ```
 
+**User responds**: "When I click the chart on backtest results, nothing happens. Console shows 'Cannot read property of undefined'"
+
+**Agent**: Now has enough context to:
+1. Search for chart-related code in backtest
+2. Look for the specific error pattern
+3. Register issue with proper details
+
 ## Important Notes
 
+- **Interview First**: Always gather context before code exploration
 - **Location Required**: Do not register without file path
 - **Evidence Required**: Do not register based on guesses without code exploration
 - **Avoid Solutions**: Do not write specific solutions (assignee decides)