The Agent Skills Directory

[SAFE]: All included Python scripts (attribution_analyzer.py, funnel_analyzer.py, campaign_roi_calculator.py) utilize only the Python standard library. No external dependencies or package managers (pip, npm) are used or required.
[SAFE]: The skill does not perform any network operations. Data processing is strictly local, reading from JSON files provided as command-line arguments and outputting analysis to stdout.
[SAFE]: No evidence of prompt injection or instructions to bypass safety guidelines was found in the documentation or script comments.
[SAFE]: The skill does not utilize dangerous functions such as eval(), exec(), or subprocess. Code execution is limited to deterministic mathematical calculations and data parsing.
[SAFE]: No obfuscation, persistence mechanisms, or privilege escalation patterns were detected. All scripts perform calculations with standard user permissions.
[SAFE]: Data ingestion is limited to parsing JSON campaign metrics. While the skill processes untrusted data (Category 8 surface), it has no high-risk capabilities (network/system access) that could be exploited via indirect prompt injection.

campaign-analytics