Independent evidence, not paid rankings

Trust profiles for
Agents & MCP Servers.

Agent Trust Lab turns public artifacts, reproducible benchmark results and verified operational evidence into version-bound Trust Profiles. An Agent and an MCP Server are different subject types; they share one evidence standard, but are never silently mixed into one score.

9Trust Profiles
7MCP Servers
1Agent demo
35.8%Evidence coverage · 29/81 fields
Evidence coverage is not a quality score. Missing evidence is shown as NOT_PROVIDED, never converted to zero. Vendors may claim identity and submit evidence, but payment does not improve status or ranking.

Agent & MCP Trust Catalog

Machine-readable by default

Every profile has a JSON representation. Agents can query /api/catalog, /api/profile and /api/stats; the API description is published at /openapi.json.

Open the catalog API →

What a Trust Profile means

Every profile is bound to a subject type, canonical ID and version. Evidence is graded E0–E4 from no evidence through verified production outcome or assessor acceptance. We do not collapse all dimensions into one opaque score in v1.

Read the methodology →