skills/afelipeg/anthropic-skills-for-enterprise-marketing-os/price-threshold-detection/Gen Agent Trust Hub
price-threshold-detection
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements standard econometric and statistical analysis techniques (CUSUM, Hansen 2000 regression) to detect price breakpoints. No malicious code, hidden instructions, or suspicious network activity was identified during the analysis. All operations align with the stated goal of market pricing analysis.- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external user-provided files, which constitutes a potential surface for indirect prompt injection. This is addressed by the skill's focus on numeric and structured data analysis rather than instruction execution.\n
- Ingestion points:
price_volume.csvandsku_data.xlsxprocessed inscripts/demand_discontinuity.py,scripts/common_thresholds.py, andscripts/regime_switch_model.py.\n - Boundary markers: Absent for file contents; data is parsed as standard CSV/Excel structures.\n
- Capability inventory: Local execution of Python scripts via shell commands (
[bash_tool]) and file output to theresults/directory.\n - Sanitization: Data is parsed into NumPy and Pandas structures where it is treated as numeric input for mathematical operations.
Audit Metadata