home / skills / phrazzld / claude-config / log-onboarding-issues
This skill audits onboarding, generates prioritized GitHub issues for findings, avoids duplicates, and formats issues for efficient triage and fixing.
npx playbooks add skill phrazzld/claude-config --skill log-onboarding-issuesReview the files below or copy the command above to add this skill to your agents.
---
name: log-onboarding-issues
description: |
Run /check-onboarding, then create GitHub issues for all findings.
Issues are created with priority labels and structured format.
Use /fix-onboarding instead if you want to fix issues immediately.
---
# /log-onboarding-issues
Run onboarding audit and create GitHub issues for all findings.
## What This Does
1. Invoke `/check-onboarding` to audit new user experience
2. Parse findings by priority (P0-P3)
3. Check existing issues to avoid duplicates
4. Create GitHub issues for each finding
**This is an issue-creator.** It creates work items, not fixes. Use `/fix-onboarding` to fix issues.
## Process
### 1. Run Primitive
Invoke `/check-onboarding` skill to get structured findings.
### 2. Check Existing Issues
```bash
gh issue list --state open --label "domain/onboarding" --limit 50
```
### 3. Create Issues
For each finding:
```bash
gh issue create \
--title "[P0] No onboarding flow - users dropped into empty app" \
--body "$(cat <<'EOF'
## Problem
After signup, users land on an empty dashboard with no guidance. They don't know what to do first.
## Impact
- High drop-off after signup
- Users never experience core value
- Support tickets asking "how do I start?"
- Lost users who never return
## Suggested Fix
Run `/fix-onboarding` or implement:
1. **First-run detection:**
```typescript
const isNewUser = !user.hasCompletedOnboarding;
if (isNewUser) redirect('/onboarding');
```
2. **Onboarding wizard:**
- Welcome screen
- First action prompt
- Success celebration
3. **Empty state with CTA:**
"Create your first X to get started"
---
Created by `/log-onboarding-issues`
EOF
)" \
--label "priority/p0,domain/onboarding,type/enhancement"
```
### 4. Issue Format
**Title:** `[P{0-3}] Onboarding issue`
**Labels:**
- `priority/p0` | `priority/p1` | `priority/p2` | `priority/p3`
- `domain/onboarding`
- `type/enhancement` | `type/bug`
**Body:**
```markdown
## Problem
What's broken in new user experience
## Impact
Effect on activation and retention
## Suggested Fix
Code snippet, pattern, or skill to run
---
Created by `/log-onboarding-issues`
```
## Priority Mapping
| Gap | Priority |
|-----|----------|
| No onboarding flow | P0 |
| Broken auth callback | P0 |
| Paywall before value | P0 |
| No empty states | P1 |
| No first-action guidance | P1 |
| Complex initial forms | P1 |
| No loading states | P1 |
| No progressive disclosure | P2 |
| No tooltips/hints | P2 |
| No tour option | P2 |
| Retention hooks missing | P3 |
## Output
After running:
```
Onboarding Issues Created:
- P0: 1 (no onboarding flow)
- P1: 4 (empty states, guidance, forms)
- P2: 3 (progressive disclosure, hints)
- P3: 2 (notifications, email capture)
Total: 10 issues created
View: gh issue list --label domain/onboarding
```
## Related
- `/check-onboarding` - The primitive (audit only)
- `/fix-onboarding` - Fix onboarding issues
- `/cro` - Conversion optimization
- `/groom` - Full backlog grooming
This skill runs an onboarding audit and creates GitHub issues for every finding, organized by priority. It surfaces gaps in new-user experience so teams can triage and assign work without manually translating audit results into tickets.
It invokes the /check-onboarding primitive to get structured findings, parses each finding into a priority (P0–P3), checks existing open issues to avoid duplicates, and then creates formatted GitHub issues for each unique finding. Issues include a clear Problem, Impact, and Suggested Fix section and are labeled by priority, domain, and type.
Does this skill fix the issues it creates?
No. This skill only creates issues. Use /fix-onboarding to apply automated fixes or implement changes.
How does it avoid creating duplicate issues?
It queries open issues with the domain/onboarding label and checks for existing similar findings before creating new tickets.