Evidence · first-party tested

It was strong on direct retrieval and multi-document policy details, but the complex multi-hop case introduced a policy contradiction by adding a 7–12 business-day EMI reversal timeline after correctly denying cancellation and refund for a customized laptop.

◐ Mixed🧾 artifact-verifiedTested Jun 16, 2026Botpress
The observation

It was strong on direct retrieval and multi-document policy details, but the complex multi-hop case introduced a policy contradiction by adding a 7–12 business-day EMI reversal timeline after correctly denying cancellation and refund for a customized laptop.

Criterion: Retrieval accuracy ·
Query this
get_evidence({
  tool: "botpress"
})
MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Real outputs, no retouching · every cell queryable via API & MCP · aidemos.com