The inclusion rule

An entry has to clear three conditions at once. First, a primary source: at least one page on the issuer's own domain that states the product structure and the redemption rule. Second, an on-chain check: at least one contract address verified with a public JSON-RPC endpoint, recorded with the endpoint used. Third, evidenced status: operating or wound-down has to rest on a primary source, not on inference.

A project that appears only in third-party aggregators is not in this directory, however large it is. That rule removes some genuinely significant projects — including at least one major tokenisation infrastructure provider — and it is a deliberate trade.

The on-chain check, and its limits

Each address is checked with eth_getCode on the chain the token is actually deployed on, and its symbol() is read back with eth_call so the ticker in this directory is the contract's own answer rather than a brand name. 56 addresses passed that check.

Three judgement calls are worth stating plainly, because each one has bitten a directory before:

• An address returning 0x on Ethereum mainnet is not automatically invalid. For an L2 deployment that is the expected result. Judging every address on mainnet would have failed most of the correct entries here.
• A contract existing is not evidence that a product still operates. One entry in this directory has its contract intact on chain, its issuer's site offline and its redemption window closed months ago. Code on chain outlives the product.
• A brand ticker is not the same thing as an on-chain symbol. Three entries differ between the two, and the on-chain value is what appears in the contract table. The differences are listed in each entry's caveats rather than quietly normalised.

What is deliberately absent

This directory does not republish aggregated protocol tables or assets-under-management figures from third-party data services. Several such services restrict redistribution in their terms, and a figure lifted from them carries an encumbrance we cannot clear. Where a size figure matters, it can be read from the public chain instead — which is why the contract addresses are published at all.

That is a licensing decision with a visible cost: some large, legitimate projects are absent because they disclose no addresses and are described only by aggregators. 5 entries carry an empty contract list for exactly this reason, and say so rather than filling the gap.

Why shutdowns are in the directory

3 of the 22 entries are marked wound-down, and one is marked unverified. These are recorded outcomes with primary-source evidence — an issuer's own wind-down notice, an on-chain supply reading, a multisig failure — not gaps in the research. A directory that only lists survivors cannot be used to judge whether a new one is safe, which is the only reason to keep a directory at all.

One of them failed for a reason worth stating on its own: a fully-backed physical commodity token lost everything to a compromised multisig. The backing was real. The key management was not.

Freshness

Every entry carries the date it was checked against primary sources. This is a snapshot, not a live feed. Nothing in this directory is investment advice, and a status of operating is not a recommendation.

Licence: CC BY 4.0. Machine-readable copies: JSON · CSV. Corrections and additions: contact.