⌖

MCP Auditor▮

governed multi-agent security audit

Point six SAFE-T security agents at a live MCP endpoint and watch them hunt vulnerabilities in real time. Every probe runs behind a governance gate — each tool call is admitted, audited, and streamed. The target's output is treated as untrusted data, never as instructions.

SAFE-T1106 Path TraversalSAFE-T1502 Credential LeakageSAFE-T1102 Tool-Description PoisoningSAFE-T1104 Excessive ScopeSAFE-T1402 Unvalidated OutboundSAFE-T-CONTROL Schema Control
Static report renderer is the required path. Thesys C1 generative rendering (UI5) slots in behind a schema-validated seam — unknown components fall back here. No raw HTML, no untrusted links.