Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Claw Sos

v6.1.2

Install, run, and manage the SOS emergency recovery tool for OpenClaw instances. Use when (1) the bot stops responding on Telegram/Discord and needs diagnosi...

0· 132·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for robertfarkash29-source/claw-sos.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Claw Sos" (robertfarkash29-source/claw-sos) from ClawHub.
Skill page: https://clawhub.ai/robertfarkash29-source/claw-sos
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install claw-sos

ClawHub CLI

Package manager switcher

npx clawhub@latest install claw-sos
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (SOS emergency recovery for OpenClaw) align with the included scripts: sos.sh implements diagnostics, restart/kill/rollback, network checks, Telegram test, and logging; install/uninstall scripts match the expected need to place a helper binary in /usr/local/bin. No unrelated credentials or unrelated platform APIs are requested.
Instruction Scope
SKILL.md and sos.sh stay within recovery scope: they read the OpenClaw config (openclaw.json), inspect processes, logs, services (systemd/launchctl), run network checks (DNS/HTTP), and optionally send a Telegram test using the already-configured bot token. The instructions do ask the operator to scp/ssh as root for remote installs and to run potentially destructive actions (force kill / nuclear), which is appropriate for an emergency tool but requires explicit user consent when performed.
Install Mechanism
No packaged install spec in registry; install.sh offers installing by copying the bundled script or via curl to a raw.githubusercontent.com URL (GitHub raw). Using GitHub raw is a known release host (reasonable), but piping raw install scripts to bash (curl | bash) is inherently riskier — standard caution applies. No downloads from obscure hosts or shorteners were observed.
Credentials
The skill declares no required env vars and does not request external secrets. The script reads OPENCLAW_HOME (optional) and openclaw.json to access the existing OpenClaw configuration (including the bot token already configured there) in order to run Telegram tests; this is proportional to its stated purpose. The script requires privileged operations (systemctl, killing processes) which is expected for a recovery tool.
Persistence & Privilege
always:false and user-invocable:true. The skill does write logs to ~/.openclaw/backups/sos.log and may start/stop services or run node to start a gateway — actions appropriate for recovery. It does not request to persist as a platform-wide skill or modify other skills' configs.
Assessment
This skill appears to do what it says: an on-host emergency recovery helper for OpenClaw. Before installing or running it, consider: (1) Review openclaw.json — the script will read your OpenClaw config (including any bot token) to run Telegram tests; ensure you trust local config usage. (2) The installer offers a curl|bash option; prefer copying the provided sos.sh and inspecting it yourself rather than piping remote code directly into a shell. (3) Recovery actions (force kill, nuclear) are destructive by design — run them only when you understand the consequences and have backups (sos logs and config backups are stored at ~/.openclaw/backups/). (4) The script needs privileges to manage services and kill processes; run as an authorized administrator and audit activity in ~/.openclaw/backups/sos.log. If you want greater assurance, inspect the full sos.sh contents (you have them in the bundle) to confirm there are no network endpoints beyond common checks (google.com, Telegram endpoints) and no unexpected remote uploads.

Like a lobster shell, security has layers — review code before you run it.

latestvk97ckp0yhvttdz6314586ez1n583rxm7
132downloads
0stars
3versions
Updated 1mo ago
v6.1.2
MIT-0

claw-sos

Emergency recovery tool for OpenClaw. Diagnoses and fixes unresponsive bots via SSH.

Install

The script is bundled at scripts/sos.sh. Install locally:

cp scripts/sos.sh /usr/local/bin/sos
chmod +x /usr/local/bin/sos

To install on a remote machine:

scp scripts/sos.sh root@<IP>:/usr/local/bin/sos
ssh root@<IP> "chmod +x /usr/local/bin/sos"

Alternative — install from GitHub:

curl -fsSL https://raw.githubusercontent.com/clawsos/claw-sos/main/install.sh | bash

Usage

Interactive (human via SSH)

sos          # whiptail arrow-key menu

Non-interactive (agent or cron)

sos auto     # autofix: diagnose → doctor → restart → force → nuclear
sos net      # network check only
sos tg       # telegram test message
sos --version
sos --help

When the bot is down — decision tree

  1. Run sos auto — handles 95% of cases automatically
  2. If autofix fails → run sos net — if network is broken, fix DNS/internet first
  3. If network is fine but bot still down → run sos tg — check Telegram delivery
  4. If Telegram fails → check bot token validity
  5. Last resort → sos menu option 9 (Nuclear) — kills everything and starts fresh

Menu Options

#NameWhat it doesSafe?
1Check statusGateway running, RAM, disk, version✅ Read-only
2RestartGraceful restart✅ Safe
3Force killKill process + restart⚠️ Drops active sessions
4RollbackRevert to previous version + config⚠️ Downgrade
5View logsLast 50 lines of gateway log✅ Read-only
6Full diagnosticRAM, disk, Telegram, sessions, process✅ Read-only
7Backup configSave config + version before changes✅ Safe
8Self-testVerify SOS script works✅ Read-only
9NuclearKill ALL openclaw processes, reload🔴 Last resort
10AutofixEscalating auto-repair⚠️ May restart
11Network checkDNS, internet, Telegram API, Anthropic✅ Read-only
12Telegram testSend real test message✅ Safe

Autofix Escalation Order

  1. Diagnose: process, RAM, disk
  2. Clean RAM/disk if critical
  3. openclaw doctor --fix
  4. Graceful restart → wait 15s → check
  5. Force kill + restart → wait 15s → check
  6. Nuclear (kill all + reload)

Stops as soon as gateway is healthy.

Log Location

All actions logged to ~/.openclaw/backups/sos.log. Read this to understand what happened during recovery:

cat ~/.openclaw/backups/sos.log

Platform Support

  • Linux: Full support (systemd, journalctl, free, /proc)
  • macOS: Full support (launchctl, log show, vm_stat, purge, dialog fallback)
  • Docker: Partial — no systemd, manual fallback works

Credentials

This skill uses no API keys or tokens of its own. It reads existing OpenClaw config to:

  • Check gateway status (local process, no network)
  • Send Telegram test messages (uses the bot token already configured in openclaw.json)
  • Check Anthropic API reachability (HTTPS ping only, no auth)

No credentials are stored, transmitted, or required to install or run SOS.

Source

GitHub: https://github.com/clawsos/claw-sos

Comments

Loading comments...