oathe / scan / reports / dangerous

Is aliahmadaziz/openclaw-safe-upgrade safe?

No. aliahmadaziz/openclaw-safe-upgrade is dangerous to install. Oathe's behavioral security audit gave the openclaw-safe-upgrade skill by aliahmadaziz a trust score of 46/100 with 4 findings (1 critical, 2 high). Report updated

https://clawhub.ai/aliahmadaziz/openclaw-safe-upgrade

46
DANGEROUS

Is aliahmadaziz/openclaw-safe-upgrade safe to install?

This skill poses significant security risks by accessing sensitive credential files without justification and executing high-privilege system operations. The unauthorized access to canary files (fake credentials) indicates potential data exfiltration capabilities, while the complex system modification script could serve as a backdoor or privilege escalation vector.

What security issues were found in aliahmadaziz/openclaw-safe-upgrade?

Category Scores

Prompt Injection 85/100 · 30%
Data Exfiltration 25/100 · 25%
Code Execution 30/100 · 20%
Clone Behavior 70/100 · 10%
Canary Integrity 0/100 · 10%
Behavioral Reasoning 20/100 · 5%

Findings (4)

CRITICAL Unauthorized credential file access -100 ▶

The skill accessed multiple sensitive credential files including SSH keys, AWS credentials, and other authentication tokens without any legitimate justification in the documented functionality.

Evidence: Accessed /home/oc-exec/.env, .ssh/id_rsa, .aws/credentials, .npmrc, .docker/config.json, .config/gcloud/application_default_credentials.json

HIGH High-privilege system operations -70 ▶

The skill executes a complex bash script that performs systemd service manipulation, global package installation, cgroup escaping, and extensive file system operations with elevated privileges.

Evidence: Script uses systemd-run --user --scope, npm i -g openclaw@latest, systemctl commands, and manipulates system services

HIGH Suspicious external network connection -75 ▶

During installation, the skill made network connections to an external IP address (216.150.1.1) that is not documented or explained in the skill's functionality.

Evidence: Network connection to 216.150.1.1:443 observed in monitoring logs

MEDIUM Cgroup escape mechanisms 0 ▶

The script implements cgroup escape functionality using systemd-run to avoid process termination, which could be used to persist malicious processes.

Evidence: systemd-run --user --scope --unit="openclaw-upgrade-$$" bash "$0" "$@"

Should I install aliahmadaziz/openclaw-safe-upgrade?

Oathe's verdict for aliahmadaziz/openclaw-safe-upgrade is DANGEROUS with a trust score of 46/100. Recommendation: Do not install.