The Agent Skills Directory

[PROMPT_INJECTION]: No override, bypass, or administrative mode triggers were detected. The technique of repeating the prompt is used as a performance optimization for specific model architectures rather than an attempt to bypass safety guidelines.- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill does not access sensitive local files or hardcoded credentials. It operates entirely on the input provided to its transformation functions without external network transmission.- [REMOTE_CODE_EXECUTION]: There are no patterns for downloading or executing scripts from remote servers. The logic relies solely on local Python standard libraries.- [COMMAND_EXECUTION]: The code does not use subprocess, os.system, or any other mechanisms to execute shell commands or interact with the operating system.- [DYNAMIC_EXECUTION]: The implementation uses static logic and standard string manipulation. It does not employ eval(), exec(), or unsafe deserialization techniques like pickle.- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided prompts by repeating them. While it acts on untrusted data, it does not interpret or parse the content in a way that would lead to unauthorized capability execution, as it lacks dangerous tool access or file-system permissions.

prompt-repetition