Evidence · first-party tested
Introduces an off-by-one table inconsistency by labeling the final extracted advertising record with line_number 9 even though the source invoice contains only 8 records.
✗ FailedLlamaParse →
A 2-page broadcast advertising invoice PDF with nested header metadata, billing/remit addresses, 8 line items, and a financial summary. Designed to test schema-driven extraction of complex tabular invoice data with page-break continuity and precise monetary fields.
The observation
Introduces an off-by-one table inconsistency by labeling the final extracted advertising record with line_number 9 even though the source invoice contains only 8 records.
Criterion: Table & Record Completeness · Scenario: Invoice PDF (group: financial-document-extraction)
Query this
get_evidence({
tool: "llamaparse",
scenario: "financial-document-extraction"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Free with attribution.
Same input, other tools
Real outputs, no retouching · every cell queryable via API & MCP · aidemos.com