The Agent Skills Directory

[SAFE]: No security issues detected. The skill follows best practices for providing domain-specific accounting logic.
[COMMAND_EXECUTION]: The skill includes a local Python script (scripts/calculate.py) for performing depreciation and disposal calculations. The script contains standard mathematical functions and self-tests using assertions. It does not perform any network operations, file system access, or use dangerous functions like eval() or exec().
[DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or network exfiltration patterns were identified.
[PROMPT_INJECTION]: The instructions are strictly focused on accounting procedures (capitalization, depreciation methods, impairment tests) and do not contain attempts to override agent safety filters or system prompts.
[INDIRECT_PROMPT_INJECTION]: The skill defines a process for reading an asset register (asset_register.xlsx). While this represents a data ingestion surface, the operations performed on the data are purely mathematical and governed by the local calculation script, posing no significant risk of command or instruction injection.

fixed-asset-accounting