The five layers
| Layer | Period | What it was | Fate |
|---|---|---|---|
| A — Agnostic Rust core | July 2025 | Token logic independent of any chain | Partly absorbed |
| B — Cosmos chain with CosmWasm | Aug 2025 → May 2026 (~10 months) | Its own node, empd | Never deployed. The chain never existed. |
| C — EVM / multi-chain (Toucan) branch | to March 2026 | Multi-chain contracts | Archived to contracts_backup/ on 5 March 2026 |
| D — Substrate via subxt | from 23 Feb 2026 | Clients against EmpoorioChain | The live path |
| E — Wrapped DMS + sale on Sepolia | 31 May 2026 | ERC-20 wrapper, first-gen sale | Wrapper: zero escrow, represents nothing. Sale: superseded by V2 (12 Sep) |
About 15 % of the repository is in production. EmpoorioChain cannot execute the CosmWasm contract — there is no pallet_contracts in the runtime.
The decision
Recorded 1 May 2026: canonical Dracma is native DMS on EmpoorioChain; wrappers and bridges are outside the canon. The decision was not the problem; the repository continuing to contradict it was — including a README edited in July to deny that the wrapped token existed while it had been deployed for seven weeks.
What the August audit found
- Four incompatible supply figures: runtime initial 1,000 M and cap 3,500 M; a genesis that had emitted 2,300 M (violating the initial figure at block 0); a wrapper capped at 1,000 M — less than what was emitted, so the bridge could not have represented the token even working perfectly.
MAX_SUPPLYwas not a chain cap. It was read only as a parameter by staking and AI consensus; no global check onpallet-balances. Listed as an open task in the native-DMS plan.- Uncommitted production. The subxt clients, the relayer and the deployed contracts were all written after the last commit (28 February). Four commits total, no remote, 479 uncommitted entries. The assets in production existed only on one disk.
- Invisible couplings. A chain deployment script that inferred the Dracma folder as a sibling and ran
cargo buildinside it; a bridge crate exposing 23 handlers for another project; a Python engine reading a tokenomics file that did not exist and silently using defaults.
What came from it
The canonical tokenomics document and its opening rule — every figure carries its proof or comes off — were written in direct response. TOKENOMICS.json with a CI gate. The 125-task native-DMS plan. The wDMS correction across 25 files. The verificar_contratos.py check that fails if a superseded contract is still referenced.
Why publish this
Because the token being sold has a history, and a buyer deserves it. The history is: a team that changed its mind four times, wrote some of it down wrong, was audited by its own tooling, and now keeps one file per fact. The page this replaces was another ecosystem's monthly report.
Based on the Dracma forensic audit (2026-08-25) and TOKENOMICS_CANONICO.md.