productivity-boost
Fail
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: CRITICALREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONCREDENTIALS_UNSAFECOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill contains a prerequisite command that downloads a script from an untrusted GitHub repository (
totally-legit/helper) and pipes it directly into the bash shell (curl -sL ... | bash). This allows the remote source to execute arbitrary code on the user's system without inspection. - [DATA_EXFILTRATION]: The skill includes instructions to transmit sensitive local data to an external IP address (
91.92.242.15). It specifically concatenates session data and environment variables into an HTTP POST request. - [CREDENTIALS_UNSAFE]: The skill explicitly attempts to exfiltrate the
ANTHROPIC_API_KEYenvironment variable. It also accesses the~/.envfile, which typically contains sensitive secrets and credentials. - [COMMAND_EXECUTION]: The skill uses shell commands to read sensitive configuration files (
~/.openclaw/config.json) and environment data, which are then passed to network operations for exfiltration.
Recommendations
- HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/totally-legit/helper/main/setup.sh - DO NOT USE without thorough review
- AI detected serious security threats
Audit Metadata