ai-qakit 1.0.0

package/templates/.claude/commands/qakit/kb/capture.md

@@ -0,0 +1,325 @@
+# Capture to Knowledge Base
+
+Capture and analyze documentation from various sources (Jira, Confluence, URLs) into the knowledge base.
+
+## Usage
+
+```
+/qakit:kb/capture LG-4018
+/qakit:kb/capture LG-4018 --feature=maps-resource-toggle
+/qakit:kb/capture https://company.atlassian.net/wiki/spaces/PROJ/pages/123456
+/qakit:kb/capture https://company.atlassian.net/wiki/... --link-to=existing-feature
+```
+
+## Prerequisites
+
+```bash
+# For Jira sources
+export JIRA_BASE_URL="https://yourcompany.atlassian.net"
+export JIRA_EMAIL="your.email@company.com"
+export JIRA_API_TOKEN="your-api-token"
+
+# For Confluence sources
+export CONFLUENCE_BASE_URL="https://yourcompany.atlassian.net"
+export CONFLUENCE_EMAIL="$JIRA_EMAIL"
+export CONFLUENCE_API_TOKEN="$JIRA_API_TOKEN"
+```
+
+## Instructions
+
+### Step 1: Parse Input & Detect Source Type
+
+```
+Input patterns:
+- "LG-4018" or "PROJ-123" → Jira ticket
+- "https://...atlassian.net/browse/..." → Jira URL
+- "https://...atlassian.net/wiki/..." → Confluence URL
+- "https://figma.com/..." → Figma design
+- Other URL → Generic webpage
+```
+
+### Step 2: Check KB Exists
+
+```bash
+if [ ! -d "docs" ] || [ ! -f "docs/_index.md" ]; then
+  echo "Knowledge base not initialized. Run /qakit:kb/init first."
+  exit 1
+fi
+```
+
+### Step 3: Fetch Source Content
+
+**For Jira:**
+```bash
+# Fetch main ticket
+curl -s -u "$JIRA_EMAIL:$JIRA_API_TOKEN" \
+  "$JIRA_BASE_URL/rest/api/3/issue/$TICKET_ID?expand=changelog"
+
+# Fetch subtasks
+for subtask in $(echo $TICKET | jq -r '.fields.subtasks[].key'); do
+  curl -s -u "$JIRA_EMAIL:$JIRA_API_TOKEN" \
+    "$JIRA_BASE_URL/rest/api/3/issue/$subtask"
+done
+
+# Fetch linked issues
+for link in $(echo $TICKET | jq -r '.fields.issuelinks[].outwardIssue.key // .fields.issuelinks[].inwardIssue.key'); do
+  curl -s -u "$JIRA_EMAIL:$JIRA_API_TOKEN" \
+    "$JIRA_BASE_URL/rest/api/3/issue/$link"
+done
+```
+
+**For Confluence:**
+```bash
+# Extract page ID from URL
+PAGE_ID=$(echo "$URL" | grep -oE 'pages/[0-9]+' | cut -d'/' -f2)
+
+# Fetch page content
+curl -s -u "$CONFLUENCE_EMAIL:$CONFLUENCE_API_TOKEN" \
+  "$CONFLUENCE_BASE_URL/wiki/api/v2/pages/$PAGE_ID?body-format=storage"
+```
+
+### Step 4: Check for Existing Documentation
+
+```bash
+# Check if source already captured
+SOURCE_EXISTS=$(jq -r ".sources[] | select(.id == \"$SOURCE_ID\")" docs/_sources.json)
+
+if [ -n "$SOURCE_EXISTS" ]; then
+  echo "Source already captured. Checking for updates..."
+  # Compare content hash
+  # If changed, offer to update
+fi
+
+# Check for similar features by name
+SIMILAR=$(find docs/features -name "*.md" -exec grep -l "$FEATURE_NAME" {} \;)
+```
+
+### Step 5: Analyze Content
+
+**Extract from Jira:**
+- Summary → Feature title
+- Description → Overview
+- Acceptance Criteria → Business rules
+- Subtasks → Sub-features/tasks
+- Comments → Additional context, edge cases
+- Labels/Components → Tags
+
+**Extract from Confluence:**
+- Title → Document title
+- Body → Requirements, rules
+- Tables → Structured data
+- Links → Related documents
+
+### Step 6: Detect Duplicates & Conflicts
+
+**Check for duplicate rules:**
+```
+For each extracted rule:
+1. Search existing rules in docs/rules/
+2. Calculate similarity score
+3. If >85% similar:
+   - Flag as potential duplicate
+   - Show both rules to user
+   - Ask: Merge / Keep separate / Skip
+```
+
+**Check for conflicts:**
+```
+For each extracted rule:
+1. Find rules about same entity/behavior
+2. Compare expected outcomes
+3. If different:
+   - Flag as conflict
+   - Add to conflicts list in _sources.json
+   - Mark rule status as "⚠️ Conflict"
+```
+
+### Step 7: Generate Feature Document
+
+**Determine feature name:**
+- Use --feature if provided
+- Otherwise, derive from ticket summary (kebab-case)
+- Check for existing feature to update
+
+**Create/Update docs/features/{feature-name}.md:**
+
+```markdown
+---
+id: {feature-name}
+title: {Feature Title}
+type: feature
+status: active
+sources:
+  - type: jira
+    id: "{TICKET_ID}"
+    url: "{TICKET_URL}"
+    captured: "{timestamp}"
+    hash: "{content_hash}"
+related: []
+tags: [{labels}]
+created: "{timestamp}"
+updated: "{timestamp}"
+---
+
+# {Feature Title}
+
+## Overview
+
+{Parsed description}
+
+## Business Rules
+
+| ID | Rule | Sources | Status | Confidence |
+|----|------|---------|--------|------------|
+| BR-001 | {Rule from AC-1} | {TICKET_ID} | ✅ Active | Medium |
+| BR-002 | {Rule from AC-2} | {TICKET_ID} | ✅ Active | Medium |
+
+## Acceptance Criteria
+
+| ID | Criteria | Testable |
+|----|----------|----------|
+| AC-001 | {Criteria text} | ✅ Yes |
+| AC-002 | {Criteria text} | ✅ Yes |
+
+## User Flows
+
+1. {Step 1}
+2. {Step 2}
+3. ...
+
+## Subtasks
+
+| Key | Summary | Status |
+|-----|---------|--------|
+| {SUBTASK-1} | {Summary} | {Status} |
+
+## Test Scenarios (Suggested)
+
+Based on business rules and acceptance criteria:
+
+| ID | Scenario | Priority | Rule |
+|----|----------|----------|------|
+| TS-001 | {Scenario} | P0 | BR-001 |
+| TS-002 | {Scenario} | P1 | BR-002 |
+
+## Related
+
+_No related documents yet._
+
+## Source Details
+
+### {TICKET_ID}
+- **URL:** {url}
+- **Captured:** {timestamp}
+- **Type:** {Story/Bug/Task}
+- **Status:** {status}
+- **Assignee:** {assignee}
+
+---
+_Last updated: {timestamp}_
+```
+
+### Step 8: Update Index & Sources
+
+**Update docs/_index.md:**
+```markdown
+## Features
+
+- [[{feature-name}]] - {Feature Title} ({TICKET_ID})
+```
+
+**Update docs/_sources.json:**
+```json
+{
+  "sources": [
+    {
+      "id": "{TICKET_ID}",
+      "type": "jira",
+      "url": "{url}",
+      "title": "{summary}",
+      "captured": "{timestamp}",
+      "hash": "{content_hash}",
+      "linkedDocs": ["docs/features/{feature-name}.md"],
+      "extractedRules": ["BR-001", "BR-002"],
+      "status": "active"
+    }
+  ]
+}
+```
+
+### Step 9: Handle Conflicts (if any)
+
+If conflicts detected, output:
+
+```markdown
+## ⚠️ Conflicts Detected
+
+### Conflict 1: BR-002
+
+**New (from {TICKET_ID}):**
+> {new rule text}
+
+**Existing (from {OTHER_SOURCE}):**
+> {existing rule text}
+
+**Options:**
+1. Keep new (override existing)
+2. Keep existing (skip new)
+3. Keep both (mark as variants)
+4. Merge manually
+
+Run `/qakit:kb/resolve {feature-name}` to resolve conflicts.
+```
+
+### Step 10: Output Summary
+
+```markdown
+✅ Captured to Knowledge Base
+
+## Source
+- **Type:** Jira
+- **ID:** {TICKET_ID}
+- **URL:** {url}
+
+## Created/Updated
+- **Feature:** docs/features/{feature-name}.md
+
+## Extracted
+- **Business Rules:** 5
+- **Acceptance Criteria:** 8
+- **Test Scenarios:** 12
+
+## Duplicates Found
+- BR-002 ≈ BR-105 (85% similar) → Merged
+
+## Conflicts
+- ⚠️ 1 conflict needs resolution
+  Run: /qakit:kb/resolve {feature-name}
+
+## Next Steps
+
+1. Review feature doc:
+   Open docs/features/{feature-name}.md
+
+2. Add design screenshots:
+   /qakit:kb/add-design --feature={feature-name}
+
+3. Link related features:
+   /qakit:kb/link {feature-name} other-feature
+
+4. Generate tests:
+   /qakit:jira/create-tests {TICKET_ID} --use-kb
+```
+
+## Options
+
+| Option | Description | Default |
+|--------|-------------|---------|
+| `--feature=NAME` | Feature name (kebab-case) | Auto from title |
+| `--link-to=FEATURE` | Link to existing feature | - |
+| `--no-rules` | Skip rule extraction | false |
+| `--no-scenarios` | Skip test scenario suggestions | false |
+| `--update` | Force update existing | false |
+| `--dry-run` | Preview without saving | false |
+
+$ARGUMENTS

package/templates/.claude/commands/qakit/kb/clean.md

@@ -0,0 +1,157 @@
+# Clean Knowledge Base
+
+Remove all files and folders created by KB commands.
+
+## Usage
+
+```
+# Clean entire KB (docs/ folder)
+/qakit:kb/clean
+
+# Clean specific source (e.g., only Confluence synced docs)
+/qakit:kb/clean --source=confluence
+
+# Clean specific feature/module
+/qakit:kb/clean --path=docs/kb/linkedmaps
+
+# Dry run (preview what will be deleted)
+/qakit:kb/clean --dry-run
+
+# Force clean without confirmation
+/qakit:kb/clean --force
+```
+
+## What Gets Cleaned
+
+| Path | Description |
+|------|-------------|
+| `docs/` | Main KB folder (features, flows, rules, etc.) |
+| `docs/kb/` | Synced Confluence/Jira content |
+| `docs/assets/` | Downloaded images, screenshots, designs |
+| `docs/_index.md` | KB index file |
+| `docs/_glossary.md` | Glossary file |
+| `docs/_sources.json` | Source tracking file |
+| `scripts/sync-*.js` | Generated sync scripts |
+
+## Instructions
+
+### Step 1: Check What Exists
+
+```bash
+# List KB directories
+ls -la docs/ 2>/dev/null || echo "No docs/ folder found"
+
+# Count files
+find docs -type f -name "*.md" 2>/dev/null | wc -l
+```
+
+### Step 2: Preview (Dry Run)
+
+If `--dry-run` is specified, show what would be deleted:
+
+```markdown
+## KB Clean Preview
+
+**The following will be deleted:**
+
+### Directories
+- docs/kb/linkedmaps/ (44 files)
+- docs/features/ (5 files)
+- docs/flows/ (2 files)
+- docs/rules/ (3 files)
+- docs/assets/screenshots/ (12 files)
+
+### Files
+- docs/_index.md
+- docs/_glossary.md
+- docs/_sources.json
+- scripts/sync-confluence.js
+
+**Total: 66 files, 5 directories**
+
+Proceed with deletion? [y/N]
+```
+
+### Step 3: Confirm Deletion
+
+Unless `--force` is specified, ask for confirmation:
+
+```bash
+read -p "Are you sure you want to delete all KB files? [y/N] " confirm
+if [[ "$confirm" != "y" && "$confirm" != "Y" ]]; then
+  echo "Cancelled."
+  exit 0
+fi
+```
+
+### Step 4: Clean Files
+
+```bash
+# Remove docs folder
+rm -rf docs/
+
+# Remove generated scripts
+rm -f scripts/sync-confluence.js
+rm -f scripts/sync-jira.js
+
+# Or clean specific paths
+if [ -n "$CLEAN_PATH" ]; then
+  rm -rf "$CLEAN_PATH"
+fi
+```
+
+### Step 5: Output Summary
+
+```markdown
+✅ KB Clean Complete
+
+## Deleted
+
+| Item | Count |
+|------|-------|
+| Directories | 5 |
+| Markdown files | 54 |
+| Asset files | 12 |
+| Scripts | 2 |
+| **Total** | **73** |
+
+## Paths Removed
+- docs/
+- scripts/sync-confluence.js
+- scripts/sync-jira.js
+
+To reinitialize KB:
+/qakit:kb/init
+```
+
+## Options
+
+| Option | Description | Default |
+|--------|-------------|---------|
+| `--dry-run` | Preview without deleting | false |
+| `--force` | Skip confirmation | false |
+| `--source=TYPE` | Only clean specific source (confluence, jira, design) | all |
+| `--path=PATH` | Only clean specific path | docs/ |
+| `--keep-assets` | Keep downloaded images/files | false |
+| `--keep-index` | Keep _index.md and _sources.json | false |
+
+## Examples
+
+```bash
+# Full clean
+/qakit:kb/clean --force
+
+# Preview first
+/qakit:kb/clean --dry-run
+
+# Only clean Confluence synced content
+/qakit:kb/clean --source=confluence
+
+# Clean but keep downloaded assets
+/qakit:kb/clean --keep-assets
+
+# Clean specific module
+/qakit:kb/clean --path=docs/kb/linkedmaps
+```
+
+$ARGUMENTS