The Agent Skills Directory

[SAFE]: The skill operates entirely locally using deterministic arithmetic logic to process foreign trade accounting data. The code is transparently implemented in Node.js and does not contain network operations, data exfiltration triggers, or obfuscated payloads.\n- [COMMAND_EXECUTION]: The skill provides a CLI wrapper (scripts/run.mjs) that allows the agent to execute internal calculation logic on provided data files. This command execution is scoped to the skill's own local scripts.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted financial data provided by the user via a text input field. While this presents an ingestion surface, the data is parsed for numerical values and handled by mathematical logic rather than being interpolated into agent prompts or executed as instructions, which prevents injection-based attacks.\n- [EXTERNAL_DOWNLOADS]: The documentation in SKILL.md includes a text snippet instructing the user on how to install a payment-related utility from an official service provider (@alipay) if they choose to purchase a separate product. This download targets a trusted service and is not part of the skill's core execution flow.

export-monthly-selfcheck-free