The Agent Skills Directory

[SAFE]: The skill provides legitimate financial engineering tools and educational content. Technical analysis of the provided Python scripts confirms they use standard, well-known libraries (pandas, scikit-learn, numpy) for statistical computing and machine learning without any malicious patterns or hidden logic.
[EXTERNAL_DOWNLOADS]: The build_features.py script fetches OHLCV data from Birdeye, which is recognized as a well-known crypto market data provider. The implementation follows security best practices by using secure HTTP requests and managing API keys through environment variables instead of hardcoding.
[SAFE]: Although the skill ingests data from an external API (Birdeye), which technically constitutes a potential indirect prompt injection surface, the risk is assessed as safe. The data consists strictly of numerical OHLCV bars used for mathematical feature computation, with no mechanism for embedded text instructions to reach the agent's control logic.
[SAFE]: No evidence of prompt injection, data exfiltration, obfuscation, or persistence mechanisms was found. The skill's metadata and instructions are consistent with its stated purpose of assisting with trading feature engineering.

feature-engineering