wps-office
Pass
Audited by Gen Agent Trust Hub on Feb 24, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes untrusted data from Excel, Word, and PPT documents to perform automated tasks. * Ingestion points: Data ingestion occurs during document structure analysis and data migration tasks, such as 'Word content generation PPT' and 'Excel data import Word'. * Boundary markers: There are no explicit instructions or delimiters defined to isolate ingested document content or to prevent the agent from executing commands embedded within the text. * Capability inventory: The skill possesses significant capabilities including opening files, saving as new files, exporting documents, and printing via tools like wps_office_save_as and wps_office_print. * Sanitization: No methods for sanitizing or validating ingested content are described before the agent processes it for downstream actions.
- [COMMAND_EXECUTION]: The skill defines and uses a set of tools to manage local application states and the file system. * Evidence: Functional tools such as wps_office_open_file and wps_office_save_as allow for direct interaction with local storage and application interfaces, which is necessary for the skill's functionality but requires strict control over the input source.
Audit Metadata