meta-spreadsheet-insight
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data from external spreadsheets and uses it to drive subsequent actions.
- Ingestion points: The
readstep inSKILL.mdextracts data from user-provided Excel workbooks. - Boundary markers: The data passed to the
analyzestep ({{ outputs.read }}) lacks delimiters, markers, or explicit instructions to the LLM to ignore any embedded commands within the spreadsheet content. - Capability inventory: The skill possesses write capabilities, including appending new sheets to workbooks (
xlsxskill) and saving information to long-term memory (memoryskill). - Sanitization: There is no evidence of sanitization, filtering, or validation performed on the spreadsheet content before it is passed to the summarization and memory-storage components.
Audit Metadata