The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references and retrieves data from official sources including the World Bank WDI API, IMF World Economic Outlook, and OECD Statistics. These are trusted global economic institutions and well-known services. The retrieval is implemented via established libraries such as wbdata and requests for legitimate analytical purposes.
[INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface where it processes external economic indicators from third-party APIs (World Bank, IMF). 1. Ingestion points: Data is fetched in scripts/fiscal_trap_analyzer.py using the wbdata library. 2. Boundary markers: The data is handled as structured numeric content (Pandas DataFrames), though no explicit natural language delimiters are used. 3. Capability inventory: The skill generates JSON data files and PNG/PDF visualization reports. 4. Sanitization: The ingested data undergoes mathematical Z-score normalization and linear regression, which effectively prevents arbitrary text from being interpreted as instructions. The risk is evaluated as safe given the quantitative context.
[COMMAND_EXECUTION]: The provided Python scripts (fiscal_trap_analyzer.py and visualize_combined.py) perform data processing and visualization. The execution is confined to standard data science operations and does not involve high-risk commands, privilege escalation, or persistence mechanisms.
[SAFE]: No malicious patterns, hardcoded credentials, or obfuscated payloads were identified. The skill's behavior matches its stated purpose of demographic and fiscal risk analysis.

demographic-fiscal-trap-analyzer