The Agent Skills Directory

[SAFE]: The skill acts as an educational and technical resource for developers. It describes legitimate data processing techniques without any malicious intent or hidden behavior.
[EXTERNAL_DOWNLOADS]: The skill references established and trusted libraries such as transformers, openai, anthropic, and unsloth. All remote operations are directed towards official API endpoints or well-known model registries (HuggingFace).
[PROMPT_INJECTION]: No attempts to override model instructions, bypass safety filters, or extract system prompts were found in the documentation or code snippets.
[DATA_EXFILTRATION]: The skill does not access sensitive system files (e.g., SSH keys, AWS credentials). Network operations are strictly limited to the intended functionality of sending prompts to official LLM providers.
[COMMAND_EXECUTION]: The provided code examples use standard Python libraries for tokenization and data analysis. There is no use of unsafe functions like os.system, subprocess.run with untrusted input, or dynamic code evaluation.

superbpe