The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data regarding income, existing debts, and employment details in Step 3 of the instructions. This ingestion of external data creates a surface for indirect prompt injection attacks where a user might attempt to override agent instructions. \n- Ingestion points: User-provided financial details collected in SKILL.md and SKILL_HE.md. \n- Boundary markers: The skill does not implement specific delimiters or safety warnings to ignore instructions embedded within the user's financial data. \n- Capability inventory: The skill utilizes Bash(python:*) for mortgage calculations and WebFetch for data retrieval. \n- Sanitization: No explicit input sanitization or validation is defined for the financial data inputs.\n- [EXTERNAL_DOWNLOADS]: The skill fetches data from https://boi.org.il/PublicApi/GetInterest. This is the official public interest rate API of the Bank of Israel, a trusted government entity, and is documented neutrally as a source for current rates.

israeli-mortgage-comparator