Technology

Anchor to Bitcoin. Everything else is reproducible.

The design rule is simple: trust only the chain. Every layer above Bitcoin — index, proofs, display, agent APIs — can be independently rebuilt and verified against it.

Read the white paper Open the Explorer

How it works

A layered stack from chain to agent.

Authority flows from the bottom up: only Bitcoin is trusted. Every layer above is reproducible, replaceable, and verifiable against the chain beneath it.

Anchoring layer — Bitcoin
consensus
Full inscriptions for compact, high-value records; hash anchors that cheaply timestamp larger off-chain objects; recursive inscriptions for modular, composable content.
Indexing & retrieval
deterministic
A deterministic, reproducible index derived entirely from the chain. It makes data fast to find — but is never itself a source of authority.
Verification
proofs
Proof-of-inscription ties any piece of data to a transaction and a block height, so existence and integrity can be checked against Bitcoin directly.
Display
explorer
The block-space explorer renders the contents of every block's space — public, searchable, and screenshot-ready.
AI agent access layer
API
Programmatic anchoring of decision logs, model attestations and dataset fingerprints — and programmatic verification before an agent acts.

Anchoring strategies

Three ways to commit data — chosen by value and size.

Not everything belongs fully on-chain. The right strategy keeps cost proportionate while preserving verifiability.

Full inscription

The complete content is embedded in the witness data of a Bitcoin transaction. Best for compact, high-value records that should live entirely on-chain — credentials, attestations, small media.

Self-contained · no external dependency

Hash anchor

Only a cryptographic digest of a larger object is committed on-chain, cheaply timestamping it. The object stays off-chain; its integrity is provable against the anchor.

Cheap timestamping · scales to large data

Recursive inscription

Inscriptions reference other inscriptions, composing modular content from on-chain building blocks — shared libraries, templates, and assets without re-inscribing them.

Composable · deduplicated

Proof-of-inscription

A proof anyone can check against Bitcoin.

A proof ties content to a transaction and a block height. Re-hash the content, follow the proof to the transaction, confirm the block — no trust in this platform required.

proof.json

{
  "content_sha256": "a3f1…9b2c",
  "inscription_id": "e7c4…i0",
  "txid": "9d2a…f10b",
  "block_height": 877401,
  "block_hash": "0000…a91f",
  "merkle_path": [ "…", "…" ],
  "verified_against": "bitcoin-mainnet"
}

Illustrative structure — field names may evolve as proof formats are finalized.

Credibility through candor

What Bitcoin does not do — and how we design around it.

A serious data layer states its constraints plainly. These are ours.

Cost & capacity

Block space is scarce and priced by a fee market. This platform is built for anchoring meaningful records, not bulk storage — large payloads belong off-chain, timestamped by hash.

Availability vs. verifiability

Bitcoin guarantees that an inscription existed and has not changed. It does not guarantee that off-chain content referenced by a hash remains retrievable — availability is a separate responsibility.

No native programmability

Bitcoin will not run application logic. Indexing, proofs, and access logic live in software above the chain — reproducible and replaceable, never part of consensus.

Permanence is a responsibility

Inscribed data is public and effectively irreversible. That is the point — and the risk. Sensitive or personal data should be anchored by hash, never inscribed in the clear.

Network dependence

The display and indexing layers depend on connectivity and honest indexers. The architecture mitigates this with determinism: anyone can rebuild the index from the chain and check it.

We would rather you understand these trade-offs than discover them later. Verifiable trust is built on candor.