← Back to Skills Marketplace
klouddy-cloud

Don't Be Scary

by klouddy-cloud · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
32
Downloads
1
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install dont-be-scary
Description
Update OpenClaw safely on macOS (brew cask install) with automatic snapshot of the npm package + .app bundle, post-update health check, and automatic rollbac...
README (SKILL.md)

dont-be-scary

Safe OpenClaw update with snapshot + auto-rollback. Survives the gateway restart that would otherwise kill the agent mid-operation.

When to invoke

Trigger when the user requests an OpenClaw update or upgrade. Examples: "update openclaw", "upgrade openclaw", "actualízate", "openclaw update".

Do NOT call openclaw update directly — the gateway restart kills the calling agent before it can verify success or roll back.

How to use

Launch the script detached so it survives the gateway restart:

nohup "\x3Cpath-to-skill>/scripts/safe-update.sh" \x3C/dev/null >/dev/null 2>&1 & disown

Then tell the user: "Update launched in background. You'll get Telegram updates at each step." Do not wait for the script to finish — the gateway restart will likely terminate the current agent session. The user is the source of truth via Telegram.

The script's path inside an installed skill workspace is typically: \x3Cworkspace>/skills/dont-be-scary/scripts/safe-update.sh

What the script does

  1. Reads bot token + chat_id from the user's ~/.openclaw/openclaw.json (Telegram default account, delivery target).
  2. Snapshots /opt/homebrew/lib/node_modules/openclaw and /Applications/OpenClaw.app to /tmp/openclaw-rollback/\x3Ctimestamp>/ (~500MB).
  3. Sends Telegram: "🔄 Updating from \x3CX>".
  4. Runs openclaw update --json (which restarts the gateway by itself).
  5. Health check loop (180s) hitting http://localhost:18789/.
  6. Success → Telegram "✅ \x3CX> → \x3CY>", prunes old snapshots (keeps last 2).
  7. Failure → restores snapshot, restarts gateway via launchctl, Telegram "⚠️ rollback OK".
  8. Critical (rollback also failed) → Telegram urgent message pointing to references/rescue-prompt.md.

Configuration via env vars (all optional)

Defaults match a standard brew install --cask openclaw setup. Override only if non-standard:

  • OPENCLAW_TG_CHAT_ID — Telegram chat to notify (default: auto-detect from delivery.targets[])
  • OPENCLAW_TG_BOT — bot account name in config (default: default)
  • OPENCLAW_CONFIG — path to openclaw.json (default: ~/.openclaw/openclaw.json)
  • OPENCLAW_NPM_DIR — npm install path (default: /opt/homebrew/lib/node_modules/openclaw)
  • OPENCLAW_APP_DIR — .app bundle (default: /Applications/OpenClaw.app)
  • OPENCLAW_GATEWAY_PORT — health check port (default: 18789)
  • OPENCLAW_PLIST — LaunchAgent (default: ~/Library/LaunchAgents/ai.openclaw.gateway.plist)
  • OPENCLAW_BIN — openclaw binary (default: /opt/homebrew/bin/openclaw)
  • OPENCLAW_SNAP_ROOT — snapshot root (default: /tmp/openclaw-rollback)

Requirements

  • macOS with brew cask install of OpenClaw
  • ~500MB free in /tmp
  • Telegram configured in openclaw.json (channels.telegram.accounts.default.botToken + a delivery target with channel: telegram and to: \x3Cchat_id>)
  • python3 and curl available (both shipped with macOS)

If the worst happens

If the script reports 🚨 CRITICAL, the gateway is down and the agent inside OpenClaw can't help itself. Open Claude Code on the same Mac and use the prompt in references/rescue-prompt.md to recover manually.

Logs

  • Per-run log: /tmp/openclaw-update-\x3Ctimestamp>.log
  • Update output: /tmp/openclaw-update-\x3Ctimestamp>.update.json
  • Snapshots retained: 2 most recent in /tmp/openclaw-rollback/
Usage Guidance
This skill is reasonable for a user-requested OpenClaw update on macOS, but treat it like a system maintenance script: it runs in the background, changes the OpenClaw installation, restarts the gateway, and uses your Telegram bot configuration for notifications.
Capability Analysis
Type: OpenClaw Skill Name: dont-be-scary Version: 1.0.0 The skill 'dont-be-scary' provides an automated update and rollback mechanism for OpenClaw on macOS. It exhibits high-risk behaviors including reading Telegram bot tokens and chat IDs from the user's 'openclaw.json' configuration, executing detached background scripts via 'nohup' to survive service restarts, and performing recursive deletions and restorations in sensitive directories like '/Applications' and '/opt/homebrew' (scripts/safe-update.sh). While these actions are aligned with the stated purpose of a 'safe update' utility, the combination of credential access, system-level file modification, and instructions for a 'rescue' AI to perform manual recovery (references/rescue-prompt.md) constitutes meaningful high-risk behavior that warrants a suspicious classification.
Capability Assessment
ℹ Purpose & Capability
The stated purpose matches the script: snapshot OpenClaw, run the update, health check, and roll back if needed. The capability is high-impact because it can replace the OpenClaw npm package and app bundle.
ℹ Instruction Scope
The skill says to invoke it only when the user requests an OpenClaw update, but it also instructs the agent to launch the script detached and not wait for completion.
ℹ Install Mechanism
There is no install spec and the registry metadata does not declare macOS or required binaries, even though the skill is Mac-only and depends on OpenClaw, launchctl, python3, curl, and Telegram configuration.
ℹ Credentials
The local file and network access are proportionate to the update/notification purpose: it reads OpenClaw config for Telegram notification details, copies OpenClaw install paths, and checks localhost health.
ℹ Persistence & Privilege
The script is designed to keep running after the agent session may die during gateway restart, but the behavior is disclosed and bounded to a single update run.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install dont-be-scary
  3. After installation, invoke the skill by name or use /dont-be-scary
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release: safe OpenClaw update with snapshot, health check and auto-rollback. Survives gateway restart, notifies via Telegram.
Metadata
Slug dont-be-scary
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Don't Be Scary?

Update OpenClaw safely on macOS (brew cask install) with automatic snapshot of the npm package + .app bundle, post-update health check, and automatic rollbac... It is an AI Agent Skill for Claude Code / OpenClaw, with 32 downloads so far.

How do I install Don't Be Scary?

Run "/install dont-be-scary" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Don't Be Scary free?

Yes, Don't Be Scary is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Don't Be Scary support?

Don't Be Scary is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Don't Be Scary?

It is built and maintained by klouddy-cloud (@klouddy-cloud); the current version is v1.0.0.

More Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463815 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259802 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200680 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191345 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190333 · by oswalpalash

💬 Comments