The Agent Skills Directory

[SAFE]: The skill implements 'Prompt Repetition', a known prompt engineering technique. The provided Python code in prompt_repetition_transformer.py performs simple string concatenation and regex-based pattern matching to determine the number of repetitions.
[SAFE]: No network operations (e.g., requests, urllib, curl) or sensitive data access patterns were identified. The skill operates entirely on the input string provided to it.
[SAFE]: The code uses only Python standard libraries (dataclasses, typing, re). It does not include any external dependencies or download remote scripts.
[SAFE]: No dynamic execution sinks such as eval(), exec(), or subprocess calls are present in the code. The wrap_llm_call function is a standard decorator pattern for transforming inputs before a function call.
[SAFE]: While the skill contains a citation to a future-dated research paper ('Google Research 2025'), this appears to be a documentation anomaly or placeholder rather than a malicious attempt to obfuscate intent. The actual behavior of the code matches its described purpose.

prompt-repetition