The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/generate_network_policy.py implements dynamic code generation and modifies file permissions.
The generate_script method accepts content from the --content command-line argument and writes it directly into a bash script file.
After creation, the script uses file_path.chmod(0o755) to make the file executable, which allows for the immediate execution of arbitrary commands if the input content is malicious.
[PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection because it processes external data while possessing high-privilege tool access.
Ingestion points: Data is ingested via the --config file path and the --content string within the scripts/generate_network_policy.py script.
Boundary markers: The skill does not utilize delimiters or specific instructions to prevent the agent from obeying commands embedded within the input data.
Capability inventory: The skill has authorized access to kubectl and the ability to write and execute scripts on the local system.
Sanitization: No validation or escaping is performed on the user-provided content before it is interpolated into shell scripts or JSON/Markdown outputs.

managing-network-policies