The Agent Skills Directory

[SAFE]: The skill serves as a reference guide for data engineering. It demonstrates standard practices such as three-stage pipeline architecture (Raw, Staging, Gold), data cleaning, validation, and splitting datasets for machine learning using standard Python libraries like sklearn, pandas, and requests.
[DATA_EXPOSURE_AND_EXFILTRATION]: While the skill includes examples that use API keys (extract_from_api), it uses placeholder variables (api_key) and follows standard practices by suggesting these be passed via headers rather than being hardcoded. No sensitive file paths or credentials are exposed.
[INDIRECT_PROMPT_INJECTION]: The skill processes external text data (web scraping, APIs). It demonstrates defensive patterns such as language detection (is_somali), quality filtering, and Unicode normalization, which are standard sanitization steps for data ingestion pipelines.
[COMMAND_EXECUTION]: There are no shell commands, subprocess calls, or dynamic execution patterns. The allowed-tools are restricted to Read and Grep, which are appropriate for a documentation and template-based skill.

etl-patterns