last30days

================================================================================

🔴 VERDICT: HIGH

This skill presents a HIGH risk due to the explicit allowance of powerful system tools (Bash, Read, Write) and the execution of a Python script (last30days.py) with user-controlled arguments ($ARGUMENTS). This combination creates a significant attack surface for command injection and data exfiltration.

Total Findings: 2

🔴 HIGH Findings: • COMMAND_EXECUTION

• Line 10, 29: The skill explicitly allows the Bash tool, which can execute arbitrary commands. Furthermore, it executes a Python script via Bash: python3 "{baseDir}/scripts/last30days.py" "$ARGUMENTS". The $ARGUMENTS are user-controlled, and if not properly sanitized by the Python script, could lead to command injection, allowing an attacker to execute arbitrary commands on the system. • DATA_EXFILTRATION
• Line 9, 10: The skill allows Read and Write tools, which grant access to the file system. Combined with the Bash tool and the implied network capabilities (via WebSearch and the skill's purpose to gather external data), there is a high potential for sensitive data to be read from the system and exfiltrated to external servers. For example, a malicious instruction could read ~/.aws/credentials and send it via a network request.

================================================================================