Bitcoin consensus confirms transactions, the Taproot script, and available witness data; the rules for tracking sats, recognizing an inscription envelope, and assigning content to a particular sat are applied by ord or another compatible indexer.
Ordinal theory numbers sats in the order they are mined and maps them from ordered transaction inputs to outputs first-in-first-out. This tracking requires no Bitcoin consensus change; it is a deterministic application-level interpretation of the same transaction chain. [2]
An inscription is created in two steps. A commit transaction creates a Taproot output committing to a script with the content, and a reveal transaction spends it through the script path, publishing the script and content in witness data. The envelope uses an unexecuted OP_FALSE OP_IF … OP_ENDIF branch and splits larger content into data pushes of at most 520 bytes. [1]
The content consists of a MIME content type and a byte string. An inscription ID is the reveal transaction txid followed by i and a zero-based index; without a pointer field, the inscription is assigned to the first sat of the relevant input. That sat then resides in an ordinary UTXO and is transferred by an ordinary Bitcoin transaction with carefully controlled ordering and values. [1]
Bitcoin full nodes validate the Taproot spend, transaction weight, and consensus rules, but Bitcoin consensus has no notion of an inscription, inscription number, or ownership of a digital artifact. An indexer reconstructs that metadata under ord rules, so indexers agree only when they use the same application-protocol version and interpretation. [3]
The reference ord project provides an index, explorer, and wallet, and its repository labels it experimental software without warranty. Its numbering covers historical edge cases such as cursed inscriptions and their vindication after Jubilee at block 824 544, so the inscription ID should be the operational identifier instead of relying on the display number alone. [4]
The witness discount lowers the relative cost of data but removes neither fees nor limits. ord documentation recommends keeping content below roughly 390 000 weight units because a standard transaction must remain below 400 000 weight units; the resulting fee changes with transaction size and fee rate, and this demand competes for the same blockspace as other transactions. [5]
Without sat control, an ordinary wallet may choose a UTXO containing an inscribed sat as an input, send it into change, or leave it to a miner as a fee. The official guide therefore separates ordinal and cardinal funds and warns that the Bitcoin Core wallet neither recognizes the inscription nor preserves the sat's position. [5]
Verification should start from the inscription ID and confirmed reveal transaction, then check the envelope, content type, exact content bytes, satpoint, and current UTXO, while distinguishing the original inscription from reinscription, parent-child links, delegation, and recursive content. A web preview or marketplace listing alone proves none of these relationships. [1]
For the clearest picture, read this entry together with Taproot, Satoshi, UTXO, SegWit, Bitcoin Script, Fee Rate. The reverse links also lead from Runes, BRC-20.
01Do Bitcoin full nodes validate Ordinals Inscriptions?+
Full nodes validate the Bitcoin transaction, Taproot script, and witness data, but they do not apply ordinal theory or assign content to a particular sat. ord or a compatible indexer does that.
02Can inscription content be changed after confirmation?+
The original inscription bytes in a confirmed reveal transaction cannot be changed. Reinscription adds another inscription to the same sat, while delegation or recursion can affect what is rendered without overwriting the original on-chain data.