Skip to content

minimum_seed_analysis.md

Latest commit

 

History

History
117 lines (84 loc) · 6.04 KB

minimum_seed_analysis.md

File metadata and controls

117 lines (84 loc) · 6.04 KB
Raw

minimum_seed_analysis.md — sBTC Receiver Seed Sizing

Author: External integrator (HK) Date: 2026-06-06 Milestone: 2 (hk-sbtc-real-receiver-v1) Method: Live mainnet reads (flashstack-sbtc-core, Velar pool 70) + integer univ2 math mirroring on-chain execution. No guessing — every constant below is read from chain. Question answered: Exactly how much real sBTC must the receiver hold before its first flash loan?

1. TL;DR

Demo loan Min seed (aggressive) Recommended Safe
1,000 sats (Matt's target) 7 sats ~50 sats ~500 sats
5,000 sats 32 sats ~100 sats ~500 sats
15,000 sats (reserve max) 97 sats ~300 sats ~1,000 sats

Plan of record: acquire ~1,000–2,000 sats of canonical sBTC, seed the receiver with ~500 sats, execute a 1,000-sat demo loan. The seed is recoverable via owner rescue-sbtc; only ~7 sats is actually consumed.

2. The repayment requirement (what the seed must cover)

Inside execute-sbtc-flash, the receiver must return owed = amount + fee to the core, or the whole tx reverts (ERR-REPAY-FAILED, and the core's own before/after reserve check). It swaps the borrowed sBTC through Velar and gets back less than it borrowed (DEX fees). The seed makes up the difference.

receiver pre-loan balance         = S            (the seed, sBTC)
core transfers borrowed amount    + L            -> holds S + L
swap L: sBTC -> wSTX -> sBTC      -> returns R    (R < L)  -> holds S + R
must repay                         owed = L + fee

solvency: S + R >= L + fee
=>  S >= (L - R) + fee  =  roundtrip_loss + fee

So minimum seed = round-trip loss + protocol fee. Nothing else touches the seed (the borrowed L is what gets swapped; the seed sits as sBTC and only tops up repayment).

3. Inputs (read from chain, 2026-06-06)

3.1 FlashStack fee — flashstack-sbtc-core

  • fee-basis-points = 5 (0.05%).
  • On-chain math: raw = amount*5/10000 (integer); fee = max(raw, 1) (1-sat floor).
  • Consequence: for any loan < 2,000 sats, raw = 0 → fee = 1 sat (the floor dominates).
Loan raw = L*5/10000 fee
1,000 0 1
2,000 1 1
5,000 2 2
10,000 5 5
15,000 7 7

3.2 Velar route loss — pool 70 (univ2-pool-v1_0_0-0070)

  • Reserves: reserve0 (wSTX) = 223,275,450,774 µSTX, reserve1 (sBTC) = 65,845,796 sats.
  • Swap fee: 0.3%/leg (get-fees → swap-fee = 9970/10000 keep ratio). Protocol takes 25% of that fee — irrelevant to the swapper's cost.
  • univ2 out (integer, as on-chain): out = (in*9970*reserveOut) / (reserveIn*10000 + in*9970).
  • Round-trip = sBTC→wSTX (leg 1) then wSTX→sBTC (leg 2), with reserves shifted by leg 1.

Because a ≤15k-sat loan is <0.025% of the 65.8M-sat reserve, price impact is negligible and the loss is essentially the two 0.3% fees → ~0.6% (60 bps) round-trip, confirmed by exact integer simulation:

Loan (sats) wSTX mid (µSTX) sBTC back round-trip loss loss (bps)
1,000 3,380,660 994 6 60.0
2,000 6,761,217 1,988 12 60.0
5,000 16,902,276 4,970 30 60.0
10,000 33,801,994 9,940 60 60.0
15,000 50,699,154 14,910 90 60.0

3.3 Slippage assumption

min-out = u1 on both legs (ADR-S4) — the swap always clears; the repayment assert + the core's reserve check are the real safety gates. There is no percentage-floor revert risk. The only "slippage" that matters economically is the deterministic 0.6% fee loss above, plus any pool move between quote and execution (covered by the recommended/safe margin).

4. Minimum seed by loan size

min_seed = roundtrip_loss + fee:

Demo loan round-trip loss core fee aggressive min recommended (≈3–7×) safe (margin for pool drift)
1,000 6 1 7 ~50 ~500
2,000 12 1 13 ~50 ~500
5,000 30 2 32 ~100 ~500
10,000 60 5 65 ~200 ~1,000
15,000 90 7 97 ~300 ~1,000

Why three tiers

  • Aggressive = exact solvency at current reserves. Leaves ~0 margin; a single block of pool movement against us could flip S + R < owed and revert (no loss of funds, just a failed tx — wasted gas).
  • Recommended = a few × the loss, absorbing realistic pool drift between the pre-flight quote and the execution block. This is what to actually seed.
  • Safe = a clean round number (~500 sats) that survives even a large adverse pool move and lets us re-run / bump the loan without re-seeding. Cost is trivial and recoverable.

5. Acquisition & total cost

  • 1 sat ≈ 223,275,450,774 / 65,845,796 ≈ 3,390 µSTX (~0.00339 STX) at pool-70 mid.
  • Seeding 500 sats costs ≈ 1.7 STX of sBTC bought on Velar (+0.3% buy fee). Seeding 1,000 sats ≈ 3.4 STX.
  • We hold 36.7 STX → ample.
  • Net economic cost of the whole demo ≈ round-trip loss (~6 sats) + 1-sat fee + buy/sell swap fees ≈ a few thousand µSTX (< $0.10-equivalent). The seed principal is recovered via rescue-sbtc and can be sold back to STX.
  • Deploy ~0.5 STX + tx fees ~0.3 STX (STX-denominated), as M1.

6. Recommendation

  1. Buy ~1,000–2,000 sats of canonical sBTC on Velar pool 70 with STX (helper: buy-sbtc.mjs, DRY_RUN first). Slight over-buy gives headroom to bump the demo loan if desired.
  2. Seed the receiver with 500 sats (safe tier for a 1,000-sat loan — survives pool drift, recoverable).
  3. Execute a 1,000-sat flash loan (Matt's stated sufficient size; well under the 15,010-sat reserve ceiling).
  4. Recover the residual seed (~493 sats) via rescue-sbtc after success.

This is the minimum-risk path that still proves the full external sBTC execution + repayment, mirroring Milestone 1.

All figures derived from live mainnet reads on 2026-06-06 and integer univ2 simulation matching on-chain arithmetic. Re-run the pre-flight quote immediately before execution — pool reserves drift, and the aggressive tier has no margin.