Paste a Relay request id, a relay.link transaction link, an explorer link, a transaction hash on any chain (EVM, Solana, Bitcoin) or a wallet. The desk finds the request, shows both legs, says what happened in plain words and writes the reply and the handoff.
Paste a ticket or chat message as it arrived. The desk reads the language, the intent and the identifiers, flags scam signs, sets a priority, lists what to ask for, and runs the lookup when a hash or wallet is present.
Every ticket is a signal, and so is the request stream behind it. This pulls the most recent requests across all users from the public API and shows what support would hear about next: the failure mix, slow fills against the quoted time, which routes and integrators carry the volume.
Before answering "is it Relay or me", check the chain. Live flags from Relay's own chain config: disabled, deposits off, block production lagging, partial limits, surge pricing.
Two kinds of errors reach support: a request's failReason (end users) and an API error body (integrators). Paste either, or browse the table.
Short replies for the cases that repeat. Each one asks for exactly what the lookup needs and never for anything that could move funds.
The part of the job that compounds: a week of tickets, clustered by cause, each cluster turned into a fix for docs, product, the developer portal or an automation. Paste your own tickets (one per line) or run the sample week.