The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/process.py is designed to be executed via a CLI to process accountability data. It performs purely mathematical calculations and data transformations without invoking shell commands or external processes.
[DATA_EXPOSURE]: The assessment script reads local JSON files provided as command-line arguments. This file access is limited to reading the input for processing and does not involve any network operations or unauthorized data exfiltration.
[INDIRECT_PROMPT_INJECTION]: A data ingestion surface is present in scripts/process.py through its processing of external JSON files. This is considered a safe implementation as the script lacks dangerous capabilities such as eval(), exec(), or network access, and it does not interpolate untrusted data into prompts.
[SAFE]: All referenced regulatory standards and enforcement precedents in references/standards.md are documented for educational purposes and do not contain executable instructions or malicious redirects.

gdpr-accountability