Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Failover Gateway Pub
v1.0.0Set up an active-passive OpenClaw failover gateway with health monitoring, auto-promotion/demotion, channel splitting, and git workspace sync for seamless re...
⭐ 0· 680·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The name/description (failover gateway) aligns with the included files and steps: a health monitor, systemd units, git sync, and promotion/demotion commands. However the health-monitor.sh contains a hardcoded default PRIMARY_IP (100.99.118.75) which is unexpected for a generic skill and could cause accidental monitoring/promotion if left unchanged.
Instruction Scope
SKILL.md instructs the operator to run remote install scripts (curl | sh for Tailscale and nvm) and to copy SSH keys and tokens; the health monitor can call git pull as another user and optionally rsync secrets from SECRETS_HOST. These actions go beyond simple orchestration and can move or pull secrets and run code from remote endpoints — acceptable for an admin guide but high-impact if done without review.
Install Mechanism
There is no automated install spec (instruction-only), which reduces automated attack surface. Still, the instructions ask you to run curl | sh against remote URLs (tailscale install and nvm install). That is a manual action but is a known risk pattern and should be audited before executing.
Credentials
The skill metadata declares no required env vars, but the guide expects channel tokens and secrets in configs and the monitor script can rsync secrets from a remote SECRETS_HOST. The ability to transfer secrets is built in but not surfaced in metadata, so sensitive credentials are in-scope for the deployment even though the registry shows none — this mismatch is misleading and risky.
Persistence & Privilege
The deployment creates persistent systemd units and places a script in /usr/local/bin: normal for a failover agent, but the service will have permission to start/stop the OpenClaw gateway and perform git/rsync operations. No 'always' privilege is requested, but the system-level persistence and sudo/systemctl operations are powerful and should be intentionally authorized.
What to consider before installing
This reads like an admin playbook rather than malicious code, but take precautions before following it:
- Do not blindly run curl | sh commands — inspect the installation scripts (Tailscale, nvm) before executing.
- Before enabling the health monitor, edit the health-monitor.sh variables: set PRIMARY_IP to your primary’s IP and clear any default addresses. Leaving the default IP will cause the monitor to probe and react to a third-party host.
- Treat SECRETS_HOST as highly sensitive: the monitor can rsync secrets from that host. Only set it to a trusted, secured host and ensure SSH keys and rsync access are limited. If you don't need remote sync, leave SECRETS_HOST empty.
- Ensure the standby’s OpenClaw channel tokens are separate from primary tokens (the guide recommends this); never copy primary channel tokens to the standby.
- Review the included health-monitor.sh and systemd unit files to confirm the commands run as the intended user and that file permissions are correct.
- Test this entire flow in an isolated staging environment before deploying to production to verify promotion/demotion behavior and avoid accidental failovers.
Given the hardcoded default IP and the optional secret-sync step, treat this skill as potentially risky until you manually audit and adapt it to your environment.Like a lobster shell, security has layers — review code before you run it.
latestvk978qn5rs965szdd5xym7x2g9s817zr1
License
MIT-0
Free to use, modify, and redistribute. No attribution required.