Diagnoses and auto-heals BlueBubbles ↔ OpenClaw iMessage connectivity. Use when: iMessages stop arriving after a gateway restart, webhook connection is broken, or user reports messages not coming through. Runs a 4-step diagnostic and auto-fixes webhook backoff, stale registrations, and gateway issues.
Scanned 9/5/2026
Install to Claude Code
npx -y skills add dvcrn/openclaw-skills-marketplace --skill bluebubbles-healthcheck --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Bluebubbles Healthcheck?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/dvcrn-bluebubbles-healthcheck)More formats (shields.io, HTML) on the badges page.
---
name: bluebubbles-healthcheck
description: "Diagnoses and auto-heals BlueBubbles ↔ OpenClaw iMessage connectivity. Use when: iMessages stop arriving after a gateway restart, webhook connection is broken, or user reports messages not coming through. Runs a 4-step diagnostic and auto-fixes webhook backoff, stale registrations, and gateway issues."
homepage: https://github.com/amzzzzzzz/bluebubbles-healthcheck
---
# BlueBubbles Healthcheck Skill
## When to Use This Skill
Use this skill when:
- iMessages aren't being delivered to/from OpenClaw
- After restarting the OpenClaw gateway
- User reports "messages not coming through"
- Periodic healthcheck (can be added to HEARTBEAT.md)
- Debugging BlueBubbles ↔ OpenClaw connectivity
## What It Does
Diagnoses and auto-heals the webhook connection between BlueBubbles and OpenClaw. This is a common failure mode: after gateway restarts, BlueBubbles can lose its webhook or enter backoff state.
**Diagnostic checks:**
1. BlueBubbles server reachable
2. Webhook registered pointing to OpenClaw
3. OpenClaw gateway endpoint responding
4. Recent webhook delivery activity
**Auto-healing:**
- Restarts OpenClaw gateway if endpoint is down
- Deletes stale webhooks and re-registers fresh
- Verifies fix after healing
## How to Use
### Quick Check (Read-Only)
```bash
BB_URL="http://127.0.0.1:1234" \
BB_PASSWORD="your-password" \
~/.openclaw/workspace/skills/bluebubbles-healthcheck/scripts/diagnose.sh
```
**Interpret the output:**
- All ✅ = healthy, no action needed
- Any ❌ = issue detected, consider running heal
### Auto-Heal
```bash
BB_URL="http://127.0.0.1:1234" \
BB_PASSWORD="your-password" \
~/.openclaw/workspace/skills/bluebubbles-healthcheck/scripts/heal.sh
```
This will:
1. Run diagnostics
2. Identify what's broken
3. Attempt to fix it (gateway restart, webhook reset)
4. Re-run diagnostics to verify
### Dry Run (See What Would Happen)
```bash
BB_URL="http://127.0.0.1:1234" \
BB_PASSWORD="your-password" \
~/.openclaw/workspace/skills/bluebubbles-healthcheck/scripts/heal.sh --dry-run
```
## Environment Variables
| Variable | Required | Default | Description |
|----------|----------|---------|-------------|
| `BB_URL` | Yes | `http://127.0.0.1:1234` | BlueBubbles server URL |
| `BB_PASSWORD` | Yes | — | BlueBubbles API password |
| `OPENCLAW_WEBHOOK_URL` | No | `http://127.0.0.1:18789/bluebubbles-webhook` | OpenClaw webhook endpoint |
You can also pass these as args: `--bb-url`, `--password`, `--webhook-url`
## Agent Decision Flow
```
User reports iMessage issue
↓
Run diagnose.sh
↓
┌────┴────┐
│ All ✅? │
└────┬────┘
Yes │ No
↓ │ ↓
Report │ Run heal.sh
healthy │ ↓
│ ┌───┴───┐
│ │Fixed? │
│ └───┬───┘
│ Yes │ No
│ ↓ │ ↓
│Report│ Escalate to user:
│fixed │ - BB app not running?
│ │ - Network issue?
└──────┴─ Manual intervention needed
```
## Common Failure Patterns
### Pattern 1: Gateway restart broke webhooks
**Symptoms:** Messages stop after `openclaw gateway restart`
**Fix:** `heal.sh` will reset webhook
### Pattern 2: BlueBubbles in backoff
**Symptoms:** Webhook exists but BB stopped trying to deliver
**Fix:** `heal.sh` deletes and re-registers webhook (clears backoff state)
### Pattern 3: Gateway not running
**Symptoms:** Check 3 fails (port 18789 not listening)
**Fix:** `heal.sh` runs `openclaw gateway restart`
### Pattern 4: BlueBubbles.app not running
**Symptoms:** Check 1 fails (HTTP 000)
**Fix:** Manual — user must start BlueBubbles.app on the Mac
## Files
```
skills/bluebubbles-healthcheck/
├── SKILL.md ← You are here
├── README.md ← GitHub docs
└── scripts/
├── diagnose.sh ← Read-only diagnostics (exit 0 = healthy)
├── heal.sh ← Auto-heal orchestrator
└── reset-webhook.sh ← Atomic webhook delete+re-register
```
## Security Notes
### Why does the webhook URL contain the password?
`reset-webhook.sh` registers a webhook URL like:
```
http://127.0.0.1:18789/bluebubbles-webhook?password=...
```
This is a **BlueBubbles → OpenClaw authentication constraint**, not arbitrary exposure. When BlueBubbles fires webhook events, it calls this URL. OpenClaw's BB plugin uses `?password=` to verify the incoming callback is from a trusted source. There is no other mechanism in the current BB↔OpenClaw integration for authenticating inbound webhook calls.
**Mitigations already in place:**
- Both services run on `127.0.0.1` (localhost only — never exposed externally)
- The password is masked in all log output by the script
- The URL is only stored inside BlueBubbles' local config (not transmitted off-device)
**What you should know before installing:**
- `BB_PASSWORD` will be stored inside BlueBubbles' webhook config on disk
- Only use on machines where both BB and OpenClaw run locally and are trusted
- Do not point `BB_URL` at a remote BlueBubbles instance
### Required binaries
| Binary | Used by | Notes |
|--------|---------|-------|
| `curl` | All scripts | HTTP calls to BB API |
| `python3` | diagnose.sh, reset-webhook.sh | JSON parsing |
| `nc` | diagnose.sh, heal.sh | Port check on 18789 |
| `openclaw` | heal.sh | Gateway restart (gracefully skipped if not found) |
All of these are standard on macOS except `openclaw` — this skill is part of the OpenClaw ecosystem and expects the `openclaw` CLI to be available.
## Adding to Heartbeat
To run periodic healthchecks, add to `HEARTBEAT.md`:
```markdown
## BlueBubbles Health
Every 4 hours, run the BlueBubbles healthcheck skill.
If any checks fail, run heal and report results.
```
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!