Lockally

      Use your API key in the examples

      Replaces lk_live_… in the code samples on this site so you can copy and run them. It is stored in this browser only and is never sent anywhere, including to Ask AI.

      It will be visible on the page. Use a test key if you are sharing your screen or taking screenshots.

      DNS troubleshooting

      Start here: is it a cache or a mistake

      Almost every “my record is not being picked up” is one of two things, and one command separates them:

      dig TXT _lockally-verify.yourcompany.com @ns1.yourprovider.com

      Querying your authoritative nameserver directly bypasses every cache.

      Result Meaning
      The record is there Your provider has it; you are waiting on a cache. Call /recheck
      The record is not there Your provider did not save what you think it did. Fix it there

      That distinction is worth making before anything else, because the two have completely different fixes and retrying only helps one of them.

      Check the domain exists at all

      dig yourcompany.com

      NXDOMAIN means the domain is not registered. There is nowhere to publish DNS and verification can never succeed. Lockally does not sell domains; buy it from a registrar first.

      This sounds too obvious to check, and it is the single most expensive failure in this area: a customer added a domain they had not yet bought, was told it was “registered” by wording we have since changed, and spent 23 minutes pressing Verify.

      The five things that go wrong with records

      The verification record at the wrong name. It goes at _lockally-verify.yourcompany.com, not at the apex.

      The token wrapped in a prefix. The value is the bare lkv-... token. Not lockally-verify=lkv-....

      Two SPF records. Only one TXT record on a domain may start v=spf1. Two is a permanent error for every receiving server. Merge the includes into one.

      The provider appending your domain twice. Many forms want the relative host, so entering _lockally-verify.yourcompany.com produces _lockally-verify.yourcompany.com.yourcompany.com. Use the host field from the records array, which is already relative. If in doubt, publish it and then dig it.

      A truncated DKIM value. The key is long. Some editors trim it, some providers split it into quoted chunks (which is correct). Compare what dig returns against dkim_public_record from the API.

      Domain errors, and what each one means

      Status Title What it means
      404 Domain not found Not on your account. Check spelling, and that you added it
      409 Domain claimed by another tenant Another account holds it. Not resolvable from the API; contact support
      409 Domain not verified Return-Path setup on an unverified domain. Verify first
      502 DNS lookup failed We could not query DNS at all. Transient; retry
      403 Upgrade required A paid-plan feature, such as a custom Return-Path, on Free
      403 Insufficient scope The key lacks domains:read or domains:write

      Note that /recheck needs domains:write, not read, because it also reconciles state. A read-only key gets 403 on it.

      The one that is not an error

      POST /verify returning 200 with "verified": false is not a failure response. The record was not found, and the current state is returned so you can poll. Read the verified field rather than the status code.

      Sending fails although the domain is verified

      Verified means the ownership token was seen. It does not mean SPF, DKIM or MX exist, because verification never checks them.

      So if a verified domain cannot send, or its mail lands in spam:

      1. POST /v1/domains/{domain}/recheck for live per-record status.
      2. Check SPF and DKIM are actually published; mail without them is widely filtered even when we accept it.
      3. Check the sending address is allowed. Verification authorises the domain, but the rules on which addresses you may use are in Sender identity.
      4. Remember verification does not inherit: mail.yourcompany.com needs its own.

      Still stuck

      Quote the request_id from the error body. It is on every response, and it is what lets support find your exact call rather than guessing from a timestamp.