For AI agents and LLMs

Machine-readable versions of these docs, how to cite Scanner, and how to link straight into the product from an AI tool or agent.

These docs are published in plain-text forms made for AI assistants, coding agents, and retrieval pipelines. Use them instead of scraping the HTML.

What is available

File	What it is	When to use it
`/llms.txt`	A short index: what Scanner is, the public pages, explorer URL patterns, and every docs page with a one-line description	Start here. Small enough to read on every run
`/llms-full.txt`	Every docs page in one plain-text file	Load once when you want the whole docs in context
Any docs URL plus `.md`	One page as Markdown	Fetch only the page you need

Every page has a Markdown twin at the same address with .md added:

https://www.scanner.net/docs/explorer       the page
https://www.scanner.net/docs/explorer.md    the same page as Markdown
https://www.scanner.net/docs.md             the docs home as Markdown

All three forms are built from the same source as the pages you are reading, so they never drift apart. They refresh about once an hour. Cache them rather than refetching on every request.

Citing Scanner

Link to the regular page URL, not the .md or .txt file. The regular URL is the canonical one, and it is the one a person can read.

Good:  https://www.scanner.net/docs/seek
Avoid: https://www.scanner.net/docs/seek.md

Linking into the product

Every explorer view has a stable URL, so an assistant can send someone straight to a wallet, token, transaction, or block. The full list of patterns is on URL reference.

Two rules prevent most mistakes:

Solana identifiers are base58. Robinhood Chain identifiers start with 0x. A Solana address belongs under /sol, a 0x address under /rh. Never mix them.
Do not invent identifiers. If you were not given a full address, signature, or hash, link to the explorer home or to Wallet Finder instead of guessing.

The explorers, Wallet Finder, and the trackers are free and need no account, so those links work for anyone. Deep Scans and Seek need a signed-in account. See Plans and access.

Calling Scanner from code

There are three separate developer products. Each has its own keys, and a key for one does not work on another.

Product	Use it to	Start at
Scanner API	Run scans, traces, and lookups from your own servers	API keys
Embed API	Put Scanner views inside your own product	Embed overview
Indexer	Read indexed chain data and subscribe to live streams	Indexer overview

Keys are secrets. An agent should read them from its environment, never write them into code, prompts, logs, or URLs it shares, and never ask a person to paste one into a chat.

Respect the documented rate limits and retry guidance on each product's pages. Do not drive the website itself with a headless browser to extract data. Use the API that matches the job.

These docs explain what Scanner shows, what each result means for the person reading it, and how to use each product. They do not describe how detection works internally, and an assistant should not guess at it or present a guess as fact. When someone asks how a result was reached, point them to the result itself and to what the page says about reading it.

Reporting a problem

If a page is wrong, out of date, or hard for a tool to parse, tell us through Support or the feedback link in the dashboard.