Recognize Hipo (hGRAM) liquid staking actions

[behrang]

Hi! I'm from the Hipo team — the liquid staking protocol on TON (treasury EQCLyZHP4Xe8fpchQz76O-_RmUhaVc_9BAoGyJrwJrcbz2eZ, jetton hGRAM, formerly hTON).

Hipo traces currently render as plain jetton Burns / generic contract calls on tonviewer and in Tonkeeper. Tonstakers-style staking actions exist for the same shape of protocol (native-coin liquid staking with a jetton receipt), and Hipo's four user-facing patterns (instant/deferred stake and unstake, incl. the bill-SBT deferred paths) map cleanly onto DepositStake/WithdrawStakeRequest/WithdrawStake.

We documented the exact trace patterns with op-codes in docs/integration.md (TL-B in contracts/schema.tlb), and since straws are usually written by your team, we've prepared a proposed implementation as a PR (linked below) with golden tests pinned to real mainnet traces — treat it as a starting point; we're happy to rework it however you prefer, or for you to take it over entirely. There's also tonkeeper/tongo#499 completing the Hipo ABI schemas independently.

🤖 Generated with Claude Code

[behrang]

One more motivation worth adding to the list: #939 fixes the wallet signing preview, not just historical display.

EmulateMessageToWallet runs the emulated trace through the same bath.FindActions / DefaultStraws pipeline as historical events, so today a Hipo deposit_coins message previews in Tonkeeper as a bare outgoing GRAM transfer — the user sees coins leaving with no indication they're staking. The mechanical cause is that Hipo mints with its own tokens_minted#5445efee op instead of the standard internal_transfer#178d4519, so JettonMintFromMasterStraw never matches the mint leg. With the straws from #939, the preview shows a DepositStake on Hipo before the user signs, the same way liquidTF deposits render.

Also a heads-up on the two skip: true goldens: Hipo is about to re-enable instant minting, so fresh instant-stake traces should start landing on mainnet soon. When one does, I'll re-pin the hipo-instant-stake case to a block within public-liteserver retention so it can be recorded without an archive node.

[behrang]

Gentle ping on this and #939 — it's been about ten days with no review activity while other PRs have been landing, so I suspect it just slipped off the radar. I understand pkg/bath straws are normally written by the team, so if you'd prefer a different approach — different straw shapes, a different action mapping, or reworking the hipo implementation enum — I'm happy to redo the PR however it fits your conventions. Also note CI on #939 may be waiting for a first-time-contributor workflow approval.