• Robinhood’s official asset registry: the issuer-published list of Stock Tokens and tokenized ETFs, the same data that builds the table on docs.robinhood.com/chain/contracts. Gives the canonical address, name, ISIN and multipliers.
• Robinhood Chain itself (chain 4663): contract code, decimals(), symbol() and uiMultiplier(), read at a confirmed block, with the block number recorded.
• Robinhood Chain documentation: the general Stock Token terms (issuer, instrument, holder rights, primary market), plus the WETH and USDG addresses.

Explorer labels and community lists are useful leads but never sufficient on their own. The full ranking is on the methodology page.

• Official registry: about every hour. Each fetch is hashed; changes become events.
• On-chain reads: a full pass over every contract runs continuously, in batches.
• Documentation pages: about every six hours. A page whose content changed is no longer used to re-confirm the claims that cite it until it is reviewed.

The status page shows when each job last succeeded and how far the indexer is behind the chain.

Each claim has a freshness window, max_age_seconds, chosen for how quickly that kind of fact can change: hours for the multiplier, days for listing facts, months for legal terms, and none for a contract’s decimals. If now − observed_at exceeds it, the claim is stale, whatever it said before.

Statuses are recomputed on every request, so the API never reports old data as fresh. The envelope’s freshness object lists the stale fields of the response.

Content hashes and block numbers

Fetched sources record a SHA-256 content hash; on-chain sources record the block number. Together with the retrieval time, this lets anyone see exactly which version of a source a value came from.