Security audit
Security checks for vulnerabilities and agentic risk
The skill is mostly a legitimate spreadsheet helper, but its recalculation script persistently installs or overwrites a LibreOffice macro in the user's profile without enough scoping or opt-in.
Review before installing. Use this only if you are comfortable with a spreadsheet skill running LibreOffice locally and modifying your LibreOffice macro profile. Prefer running it in a disposable user profile or sandbox, and keep backups of important spreadsheets before allowing structural edits or recalculation.
The skill description presents a broad, general-purpose spreadsheet capability, but the documented behavior includes installing a LibreOffice macro and modifying local user configuration for formula recalculation. That side effect is materially different from ordinary spreadsheet editing/analysis and can mislead users or downstream policy systems into permitting system modification they did not expect.
The skill instructs use of file read/write operations and shell execution (python recalc.py ...) but does not declare any explicit tool scope or allowed-tools boundary. This creates unnecessary ambiguity around what capabilities the agent may exercise, increasing the risk of over-privileged execution or accidental use of broader filesystem/shell access than intended.
Subtle instructions detected that may alter agent decision-making or introduce hidden biases.
Always use Excel formulas instead of calculating values in Python and hardcoding them. This ensures the spreadsheet remains dynamic and updateable.
The examples demonstrate destructive workbook operations such as inserting rows, deleting columns, and overwriting cells without any warning, backup guidance, or confirmation requirement. In the context of existing spreadsheets, these actions can silently corrupt formulas, references, or business-critical templates if executed on the wrong file or without user intent confirmation.
subprocess module calls execute external commands. Without careful input validation, this enables command injection.
return True
if not os.path.exists(macro_dir):
subprocess.run(['soffice', '--headless', '--terminate_after_init'],
capture_output=True, timeout=10)
os.makedirs(macro_dir, exist_ok=True)
The script silently installs or overwrites a LibreOffice application macro in the user's profile, creating persistent code in a trusted macro location without notice or consent. In a spreadsheet-processing skill that may be triggered on untrusted content, this is more dangerous because it alters the host application's execution environment and can affect future LibreOffice sessions or other documents beyond the immediate task.
subprocess module calls execute external commands. Without careful input validation, this enables command injection.
if platform.system() == 'Darwin':
# Check if gtimeout is available on macOS
try:
subprocess.run(['gtimeout', '--version'], capture_output=True, timeout=1, check=False)
timeout_cmd = 'gtimeout'
except (FileNotFoundError, subprocess.TimeoutExpired):
pass
subprocess module calls execute external commands. Without careful input validation, this enables command injection.
if timeout_cmd:
cmd = [timeout_cmd, str(timeout)] + cmd
result = subprocess.run(cmd, capture_output=True, text=True)
if result.returncode != 0 and result.returncode != 124: # 124 is timeout exit code
error_msg = result.stderr or 'Unknown error during recalculation'
No suspicious patterns detected.