Coming soon

Verify MCP for AI Agents

A public trust interface that helps AI agents tell real products from unverified listings before they recommend, rank, or act.

Public verification surface

MCP online

Agent finds product

A shopping flow sees a product URL or listing.

Fetch certificate

Verify MCP exposes the public MC4P certificate (Mintall Certificate for Products).

Verify proof

JWT, JWKS, domain, and C2PA checks are validated.

Check context

The selling domain is matched to the certificate.

Return decision

The agent gets a recommendation-safe trust result.

MCP read

tools.call("mintall.verify_product")
product_url: "https://acme.com/product/hero-jacket"
returns: MC4P certificate + trust decision

Trust result

Product identityMC4P JSON-LD

Signed proofJWT + JWKS

Image provenanceC2PA present

Selling contextAuthorized domain

ELIGIBLE_FOR_RECOMMENDATION

Make product trust machine-readable

Verify MCP gives AI agents a public way to check product identity, seller context, and media provenance.

No merchant account access. No private store permissions. Just a clear trust answer an agent can use before it recommends a product.

Verify MCP verifies

Product identity

MC4P JSON-LD

Signed proof

JWT + JWKS

Image provenance

C2PA present

Selling context

Authorized domain

Verify MCP gives agent builders a simple trust primitive for commerce recommendations.

Rank verified products higher

Use public trust decisions as a ranking signal before an agent recommends a product.

Filter risky listings

Treat missing certificates, invalid proofs, or unauthorized domains as reasons to ask for more evidence.

Show proof to users

Surface certificate status, image provenance, and seller context in your shopping experience.

Call Verify MCP or the Verification API from your agent and get a trust decision for any product before you recommend it.

Building a marketplace or platform? See Mintall Enterprise