Skip to content

CoreWriter actions succeed on EVM but produce no effect on HyperCore (testnet) #172

Description

@wsim101

Summary

CoreWriter (0x3333...3333) actions succeed on HyperEVM testnet (tx status 1, RawAction event emitted with correct data) but produce no effect on HyperCore whatsoever — no fills, no ledger entries, nothing. Confirmed across order placement (buy and sell) and plain token transfers.

Environment

Reproduction

0xaFD5292Cb3B71858c82e930C7D9902e1D2a5B59A — freshly created, funded, and bridged in specifically for this test, no prior history

Both accounts, on both days tested (2026-09-20 and 2026-09-21), sent well-formed CoreWriter actions:

  • Spot limit orders (id 1) — buy and sell, HYPE/USDC market (asset id 11035), IOC, permissive limit prices (10-20% through the opposing best price), $10+ notional, sizes rounded to the required 0.01 HYPE lot
  • Spot send (id 6) — a simple self-transfer

All succeeded on EVM (status 1, correctly-decoded RawAction event) and all produced zero HyperCore-side effect: no userFills entry, no userNonFundingLedgerUpdates entry, no balance change, no open order, no error of any kind.

Example tx: 0x0c552424236c246e67bd7acbe4c54fb4901d501dd8610ef3e3f7257d9891d228 (buy, block 64825883).

Ruled out

  • Encoding — action bytes decoded and cross-checked against the documented layout (uint32,bool,uint64,uint64,bool,uint8,uint128) — correct
  • Notional/size — properly sized, above the $10 minimum, correct lot rounding
  • Slippage — limit prices set well past the opposing best price
  • Account-specific issue — reproduced identically on a brand-new, never-before-used wallet
  • Chain halt — HyperCore's own block counter advances normally (~20+ blocks/sec) during testing
  • Stale/phantom liquidity — verified live by placing a real buy through Hyperliquid's own UI, which filled cleanly against the same book a CoreWriter order had just failed against

What does work

Native EVM↔Core bridging (plain value transfer to 0x2222...2222, not a CoreWriter action) works reliably on both accounts — funds land in HyperCore spot balance as expected, activation fee charged normally. The failure is specific to actions routed through sendRawAction.

Expected behavior

A well-formed CoreWriter action should be relayed to and processed by HyperCore, the same way it reliably did earlier in testing (both order fills and transfers worked as recently as our first testing session, 2026-09-20 early).

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions