Research Papertrade (docs.papertrade.xyz, read every page, plus papertrade.xyz, public contracts on HyperEVM, relayer or API docs) and find the most effective ways to farm PAPER, with a working solution that Swarm Pepe / OG holders can actually use. Use skill research-report. Output: artifacts/papertrade-farming-research.md (text/markdown). Research only: no trading, no deploys, no keys.

1. MECHANICS (short, exact): BBO mid pricing, 0.2 bps deadband, asymmetric impact formula and launch params (baseRate 10%, rateMultiplier 15000, referenceNotional $100k, positionMultiplier BTC 814.598 / ETH 483.979), 2% win fee, loss-basis and liquidation mint rules, PAPER mint curve (100 per $1 below $2M tracked LP, then 100·(S/(S+H))² with S = $120M, one-way ratchet), payout queue and sideBucket, staker revenue (1% of PnL, half of fees, LP excess above $5M), OI caps, min margin, session keys, relayer, deposit routes and fees.

2. FARMING STRATEGIES: search for every effective way to earn PAPER or USDC, including non-obvious ones, and rank them by cost per PAPER and risk. At minimum: hedged pair (long + short same size) vs move size, leverage, BTC vs ETH, close order, liquidation vs voluntary close; directional trading with the mint as rebate; timing (queue empty vs active, early flat-rate window, volatility); early staking rebate at 1%, 5%, 20% of total stake. Tables for moves 0.05% to 3% and leverage 10x to 1000x, formulas and worked examples.

3. WORKING SOLUTION FOR HOLDERS: design something real where a holder sends ETH and farms, not just a calculator. Compare at least: a) Non-custodial guided farming: the holder keeps their own Papertrade account; our site handles deposit (ETH to USDC to Papertrade), session key in the browser, one-click hedged pair, close-order automation and stake. Check if a third-party frontend can submit through the Papertrade relayer and what is needed.

b) Pooled vault: holders deposit ETH into a contract, the vault farms as one Papertrade account, stakes the PAPER and pays USDC/ETH back pro-rata (or to activated Pepes). Check if a smart contract can be a Papertrade account (EIP-1271 signatures, session keys, relayer rules, CREATE2 deposit proxies), the bridge path Ethereum to HyperEVM and back, and the fact that PAPER is non-transferable at launch.

c) Builder codes (planned 1% revenue share for frontends): how revenue could flow to activated Swarm Pepes through the OGDistributor 0xd450ea80aec46b8bffdf0c4f44d3964489b613f2.

For each: does it work now / at launch / never, exact steps and contracts needed, who holds funds, fees and bridge costs, expected cost per PAPER for a holder depositing 0.1, 1 and 10 ETH, risks, build effort. Recommend one to build first and write its spec so it can be built in a follow-up job.

4. RISKS: queue lock while LP is near zero, upgradeable proxies and owner powers, BBO manipulation, non-transferable PAPER, anti-sybil or wash rules, relayer dependence, custody risk for pooled funds.

5. LIVE STATE: if contract addresses are public, read tracked LP, queue length, PAPER supply and total staked from HyperEVM with the block number.

End with a one-screen TL;DR: best strategy, cost per PAPER, recommended holder solution, top 3 risks.