Build trust with your customers and their AI agents.
Free forever.
Trust, made machine-readable
TrustMCP standardizes access to your evidence, never the verdict. Drop a single /.well-known/trustmcp.json on your domain and any agent that resolves it requests a scoped key, then reaches its own conclusion.
{
"schema_version": "0.1",
"vendor_id": "vnd_acme",
"legal_name": "Acme Corp",
"network": "https://network.trustmcp.app",
"manifest": ".../v1/vendors/vnd_acme/manifest",
"mark": "agent-ready"
}An agent that knows your domain resolves from here, requests a scoped key, and reads your profile.
Upload SOC 2, pentests, ISO, COI & SBOMs and declare machine-readable claims. No more re-answering the same questionnaire.
Every customer requests access; you approve, scope, and revoke on your own terms. Every read is logged.
Any assessment agent reads your profile over MCP in a handful of tool calls, then reaches its own conclusion.
Why this works
Older assessment networks shipped a single, one-size rating. TrustMCP shares the raw, current evidence instead and lets each customer reach their own answer. Everyone keeps their own standard; we just standardize access.
Free forever for any vendor publishing evidence.
One domain verification confirms you own where you publish.
Reusable across every customer and agent.
Native to the agent tooling protocol teams already use.
Free forever, open for everyone. Publish once and let the network do the reading.