Pick a block. The Doge Block Manifest reads the whole block and sorts every transaction into the same kinds as the shelves in Doge Cube's halls. The list is proved against the block's own header, and every transaction that carries something is read from its own bytes. Each comes with its fee and size, checked on two sites. It's free, for people and AI agents, and you can download it. The Bitcoin twin is the Block Manifest.
| # | Transaction | Fee (DOGE) | Size (bytes) | DOGE/kB | Kinds | Carries | Checked |
|---|
Checked: ✓ agrees, ✗ differs, · nothing to check. The six marks are: in the list the header proves, its size, its fee, its kinds, its inscriptions, and read from its own bytes. Text in quotes like these was written by whoever made the transaction. It's shown as they wrote it, unchecked.
The whole manifest, every transaction, with its checks and where each fact came from.
Open Doge Cube and type the block number into its tablet to walk into this block's hall, where these kinds are its shelves.
Open Doge CubeIn this page: await window.agent.help(), then agent.read({height}), agent.items({kind, from, limit}), agent.item({n}), agent.checked() and agent.see(). Where the browser has WebMCP, the same commands are tools named dogemanifest_*. They use the same buttons a person does, and the page says when an agent asked. Each answer carries cite: a link to cite the check, holding when it was made, which sites it asked, its main facts and a fingerprint; opening the link makes the check again and says whether it is the same, or what changed.
The bigger apps: Doge Cube, where you walk into any Dogecoin block in 3D (its kit's walkIn({height})), and its Bitcoin twin Agent Blockhall, with the Bitcoin Block Manifest. For any Doginal in a block, the Doge Provenance Check proves where it came from, read from the chain.
Without a browser: the free Agent Blockhall MCP tool has the same Doge Block Manifest as its dogeBlockManifest tool. The block reader is one file, doge-manifest.js. More in llms.txt.