univer-pro-integrate
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill integrates features for importing and processing external XLSX and DOCX files, as well as loading data from remote collaboration servers. This represents a vulnerability surface where maliciously crafted data could be used to perform indirect prompt injection attacks against the agent.
- Ingestion points: Data enters the system via
univerAPI.importXLSXToUnitIdAsync,univerAPI.importDOCXToUnitIdAsync, anduniverAPI.loadServerUnitas documented inSKILL.md,references/exchange-guide.md, andreferences/collaboration-guide.md. - Boundary markers: The instructions do not define specific delimiters or warnings to the agent to ignore instructions embedded within the processed spreadsheet or document data.
- Capability inventory: The skill possesses network capabilities (interaction with exchange and collaboration endpoints) and file system/output capabilities (file download utilities and printing APIs).
- Sanitization: There is no evidence of sanitization or validation logic for the content parsed from these external file formats or remote units.
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install a variety of packages from the
@univerjs-proand@univerjsscopes via standard package managers. These are official vendor packages associated with the Univer Pro product described in the skill.
Audit Metadata