The Agent Skills Directory

[SAFE]: No security issues were detected. The skill provides legitimate engineering templates for model fine-tuning and optimization.- [COMMAND_EXECUTION]: The skill includes code in references/deployment-optimization.md that uses subprocess.run to call external tools for model format conversion (convert_hf_to_gguf.py) and quantization (llama-quantize). These operations are local, standard in the llama.cpp workflow, and essential for the skill's stated purpose of model optimization.- [EXTERNAL_DOWNLOADS]: The provided scripts reference several well-known and trusted machine learning libraries and services, including Hugging Face's Transformers and Evaluate libraries, Weights & Biases for logging, and OpenAI for evaluation. These dependencies are standard in the industry and originate from trusted vendors.- [PROMPT_INJECTION]: The skill facilitates the processing of external training datasets (as seen in references/dataset-preparation.md). While this constitutes an ingestion point for untrusted data, the skill includes robust validation and quality filtering logic (e.g., deduplication and regex-based content filtering) to mitigate typical risks associated with data processing.

fine-tuning-expert