The Agent Skills Directory

[SAFE]: The skill is designed for interacting with the Jinkō platform through its official Python SDK (jinko). All functionality is consistent with its stated purpose of managing data tables.
[DATA_EXPOSURE]: The scripts utilize python-dotenv to securely load environment variables (load_dotenv()). This is a recommended security practice for managing API keys and secrets in local environments, preventing hardcoded credentials.
[COMMAND_EXECUTION]: The provided scripts (scripts/create_data_table.py and scripts/inspect_data_table.py) are standalone CLI tools that perform data validation and SDK operations. They do not execute arbitrary shell commands, use sudo, or perform any unauthorized system modifications.
[EXTERNAL_DOWNLOADS]: The skill relies on standard, verifiable Python packages including jinko, python-dotenv, and pandas. No remote scripts are downloaded or executed from untrusted sources.
[INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it processes external CSV and SQLite files. However, it uses standard libraries (csv, pandas) for parsing and primarily performs validation or metadata inspection. It does not interpolate untrusted file content into the agent's prompt context in an exploitable manner.

jk-data-table