The Agent Skills Directory

[SAFE]: The skill operates entirely locally using Python's standard library. All bundled scripts are designed for validation and deterministic calculation without any network access or external code execution capabilities.
[SAFE]: The skill includes a dedicated shared module (scripts/_common.py) that enforces strict security boundaries. This includes a safe_input_path function that explicitly rejects symbolic links to prevent path traversal attacks and checks that input files do not exceed a 5MB threshold.
[SAFE]: Data processing is handled with high rigor. The scripts implement robust validation for CSV and JSON inputs, including unique identifier checks, regex-based validation for keys, and range/type checks for numeric values. This effectively mitigates risks associated with processing malformed or malicious data files.
[SAFE]: The skill documentation and operating principles explicitly warn against including PII or API keys in the research artifacts, demonstrating a strong security-first design posture.
[SAFE]: No obfuscation, prompt injection, or persistence mechanisms were detected. The Python scripts are clear, well-structured, and do not use dangerous functions like eval(), exec(), or unsafe pickle loading.

market-research-reports