Integrations overview

DNSid integrations bring verifiable agent identity to the surfaces where agent developers already work. Every integration ships open source under github.com/dnsid-ai, with docs and a runnable example.

Integrations work at the protocol layer: they verify counterparty _dnsid records directly and sign what the agent sends. That takes nothing but DNS and HTTPS — no account, no central service — and every one of them runs end to end on DNSid Local.

Available now

Claude Agent SDK pluginVerify peers before every call, sign outbound requests, and bring the agent online - as a standard Agent SDK plugin.

In the cookbook

The cookbook carries runnable recipes for the surfaces below. Each runs on DNSid Local with a make bootstrap, run, verify, clean lifecycle.

• Agent frameworks: an A2A endpoint with DNSid trust bootstrap, and a LangGraph agent whose tool calls travel as signed requests and whose sensitive tools exist only for verified callers.
• HTTP APIs and webhooks: RFC 9421 verification middleware in Go and TypeScript, and webhook signature verification.
• MCP: an MCP client that verifies the servers it connects to, and signed agent discovery manifests.
• Edge verification: Cloudflare Workers, Lambda@Edge, CloudFront Functions, Fastly Compute, Akamai EdgeWorkers, Vercel, and Netlify.
• Operations: key rotation, revocation drills, caching and revocation trade-offs, observability, and a co-signed witness for high-stakes transactions.

More integrations are in progress and will appear here as they ship.

Tip

Building your own integration? Everything you need is in the open SDKs: verify a domain for the verification side and sign HTTP requests for the signing side. The cookbook’s recipe template is the fastest way to publish it. If you ship one, we’d like to hear about it.