The Agent Skills Directory

[COMMAND_EXECUTION] (HIGH): The script tools/render-with-params.sh is vulnerable to command injection because it interpolates a shell variable directly into a Python command string.
Evidence: In tools/render-with-params.sh (line 44), the script executes python3 -c "... open('$PARAMS_FILE') as f: ...". Because $PARAMS_FILE is inside a double-quoted Bash string, the shell expands it before passing the argument to Python. An attacker using a filename containing a single quote and Python commands can achieve injection.
[REMOTE_CODE_EXECUTION] (HIGH): The command injection vulnerability in the rendering tool allows for arbitrary code execution in the context of the user running the agent.
[SAFE] (INFO): The extract-params.sh script correctly handles input by passing the filename as a command-line argument (sys.argv[1]) rather than interpolating it into the code string.
[SAFE] (INFO): The skill instructions and other utility scripts do not contain malicious patterns and appear to serve their stated purpose of CAD model management.

openscad