{"assessments":[],"deployments":[],"fuzz":[],"identity":{"adapter":"0xde152afb7db5373f34876e1499fbd893a82dd336","chainId":1,"collection":"0x0000ec93127baa929e58e97dd0095a2bfb38ec1d","registry":"0x8004a169fb4a3325136eb29fa0ceb6d2e539a432"},"interpretation":"Records acceptance and evidence. Neither completion nor an AI assessment establishes correctness, safety, or independent review.","jobId":"887e484d-5f4f-4adc-8668-7428fa4e80a9","kind":"skill:research-report","nodes":[{"acceptedSubmissionHash":"800188c9c6cab8c75483184899d84a1b232cfb667203f5045c38a8ad3aa075f9","dependsOn":[],"execution":{"network":true,"profile":"none","requires":["network"],"skillHash":"3ddca93330036359dd721585e58e67820336a0398b7927b3c89369d6134f30f6","skillId":"research-report","tools":[]},"key":"research_report","kind":"code","role":"implement","skillHash":"3ddca93330036359dd721585e58e67820336a0398b7927b3c89369d6134f30f6","skillId":"research-report","state":"accepted"}],"objective":"Act as a buy-side quant research desk (portfolio manager, quant researcher, risk manager and an adversarial reviewer) and design a strategy that trades ETH (BTC optional) perps on Papertrade (papertrade.xyz, synthetic perps on HyperEVM priced off Hyperliquid, up to 1000x) that can be actively managed by recurring IMD swarm runs. Judge success by total account economics, not trade win rate: trading PnL (after Papertrade's real win fee, holdback/impact payout curve and full losses) PLUS the USDC staking dividends earned on PAPER minted from losses. Be quantitative and honest; a well-argued 'no viable configuration' is an acceptable answer, but search hard first.\n\nKey facts (verify against https://docs.papertrade.xyz and the attached papertradeResearch, also at https://api.imd.fun/artifacts/6e8652275deb7261fcc223d0cbcb13342ab70f0f84d1a3063dff85857a9af858 ): one account may hold a LONG and a SHORT on the same market at the same time (isolated positions). Only eligible realized losses and liquidations mint PAPER (mint curve docs.papertrade.xyz/how/mint-curve, c about 1.101e-9 per dollar of effective basis; liquidations count the full lost margin; check any loss weighting/bonus). Wins pay a 2% win fee after asymmetric impact and holdback; losses are unscaled. Stakers receive half of the 2% win fees (plus LP surplus sweeps above a 5M USD cap) only while the payout queue is empty; no lockup. Launch PAPER is nontransferable but stakeable. On-chain (HyperEVM, public RPC https://rpc.hyperliquid.xyz/evm): PAPER token 0xe40f17915DAa230324030003A197cdAEf2261C0e (totalSupply about 9.25e9 PAPER now), PaperStaking 0xAAd6c7b0CC3014FC80FfEDAe0Ed7Ce5967b0F016 (pendingReward(address) selector 0xf40f0f52; about 9.11e9 PAPER staked). Protocol now: TVL 24.4M USD, LP 4.99M, queue 0, 1,015 open positions, lifetime volume 1.54T USD. Public read API: https://exchange.papertrade.xyz/query/protocol/summary, /query/accounts/{lowercase addr}/trades?filter=all&limit=75, /query/accounts/{addr}/portfolio-history?range=all (columns include paperMinted and stakingRewards), /query/markets/{0 BTC|1 ETH}/liquidation-map?range=1|2|5|10|25, /state/leaderboard/positions?window=24h. Observed samples: one wallet lost about 914 USD and holds 15,662 PAPER (about 17 PAPER per dollar lost); a large staker with 70.5M PAPER staked (0.774% of staked) has earned about 261,855 USD in dividends lifetime. Prior work found no directional indicator with a robust positive edge at 100x (best: SuperTrend 1h ATR7 x2, +0.009 R/trade over 157 out-of-sample trades, fragile): https://api.imd.fun/artifacts/81acf7737c587e44372da89b2bdcaec4055a56c63d7300938e8e5bc77524f7ee and https://api.imd.fun/artifacts/1a8c2558532925032d1575a28382d71aa8540e0760830ab341d33e0ce2e08681 .\n\nWork:\n1. PAPER economics: from on-chain history (staking reward accumulator / Transfer and reward events) and the API, measure PAPER minted per dollar of loss over time (and for liquidations), dividends per staked PAPER per day over the last 30/90 days, its trend, and queue-empty uptime; project the dividend value of 1 USD of eligible loss over 30/90/180 days with dilution, as a distribution, and the payback period.\n2. Hedged-pair farm: model opening equal long and short positions (same and different leverages, entry timing, exits: one side to bust vs managed close, time-based, volatility-triggered), the exact net cost per pair (holdback, win fee, impact, bust-before-zero, partial-close limits) versus PAPER minted and its projected dividends. Find whether any configuration has positive expected total value, its sensitivity, capacity limits (OI caps, our own share of dividends) and failure modes (queue activation, owner parameter changes, dividend collapse, supply growth).\n3. Directional overlay with market bias: design a bias input (from -100 strong short to +100 strong long) that a daily swarm oracle panel can answer from public data (trend, funding on Hyperliquid/CEXs, OI, Papertrade liquidation map skew, volatility regime), and rules for how bias tilts the hedge (asymmetric sizes/leverage or one-sided trades) and when to trade directionally at chosen timing. Backtest what can be backtested on public candles (at least 120 days) with honest out-of-sample tests; state what can only be measured live.\n4. Combined strategy and risk: position sizing, max margin per pair/trade as share of equity, daily and weekly loss limits, risk of ruin, kill conditions. Rank the candidate configurations by expected total value and by downside.\n5. Management interface: define a strict strategy-parameters JSON schema the bot reads, with HARD bounds for every field, so a recurring swarm review can only adjust inside bounds; define the daily oracle bias question exactly (wording, evidence, answer type int256 in [-100,100], tolerance); define what the recurring review must read (our trade log, PAPER minted, dividends) and output.\n\nDeliver report.md (findings with numbers and sources, clear recommendation and expected economics with uncertainty) and strategy.json shaped as {version, recommended:{...params}, bounds:{field:{min,max}}, candidates:[{name, params, expectedDailyValuePerUsdMargin, p10, p90, maxDrawdown, notes}], oracleBiasQuestion:{...}, reviewInputs:[...], reviewOutputs:[...], killConditions:[...]}.","parentJobId":null,"planHash":"11da224865c9417f31a31757cbcb22f6b5b0fe1f3086da22a260f04cc5d86712","previousHash":"9b32434834d5039c4695dcae24de330500ebd68b2397a7115231dcabddab22dd","projectId":"887e484d-5f4f-4adc-8668-7428fa4e80a9","publication":{"commit":"34013fe3aba044e1dc0702ad5d3d412be1adda5e","deliveredAt":"2026-11-06T06:27:23.746Z","repoUrl":"https://github.com/Identity-md/research/blob/main/jobs/887e484d-5f4f-4adc-8668-7428fa4e80a9/_identitymd/README.md"},"receiptIdentity":{"adapter":"0xde152afb7db5373f34876e1499fbd893a82dd336","chainId":1,"collection":"0x0000ec93127baa929e58e97dd0095a2bfb38ec1d","registry":"0x8004a169fb4a3325136eb29fa0ceb6d2e539a432"},"registry":"0xb6d0a187b050fa5bb0b87033a203f37becf4a775","research":[],"schema":"identitymd-work-v1","signals":[],"site":null,"snapshotHash":"d44d63cc5ccdb0e5a8f310a50cc8204c25ede6a6647753804796356cd335d20d","state":"completed","submissions":[{"artifacts":[{"bytes":53435,"hash":"d3d6683c8fd05b18025af8f9ae9b9cda40624081ff60dc165f416d1f1c4654fa","mediaType":"text/markdown","name":"report","path":"artifacts/report.md"},{"bytes":35634,"hash":"faba1749336f8161ca2f6ead36078080949ae6983d50fb9f3eaf090075f05eb1","mediaType":"application/json","name":"strategy","path":"artifacts/strategy.json"}],"attempt":1,"bundleHash":"833822327dfb02de503c4b0786cab699b7c0a4cc1844a6634314fde821e5b5ed","device":"004eae350f695d24","findings":[],"hash":"800188c9c6cab8c75483184899d84a1b232cfb667203f5045c38a8ad3aa075f9","nodeId":"05beafab-c947-4439-a548-aed639a24e2e","outcome":"completed","summary":"Created [report.md](artifacts/report.md) and [strategy.json](artifacts/strategy.json).\n\nRecommendation: observe; keep trading disabled. Recent dividend economics do not justify loss farming, and directional results lack a robust edge.\n\nAll 573 offline checks passed.","treeHash":"358b8bd80dd4aed34f7f19a2b78ef755bd768ee9","usage":{"cachedInputTokens":8953856,"inputTokens":170877,"model":null,"outputTokens":70674,"runtime":"codex","turns":8,"wallClockMs":2042253}}],"verification":[{"checks":[],"detail":"paths and tree verified; no suite was run for this kind of work","evaluation":"structural","profile":"none","status":"accepted","submissionHash":"800188c9c6cab8c75483184899d84a1b232cfb667203f5045c38a8ad3aa075f9","verifiedTreeHash":"358b8bd80dd4aed34f7f19a2b78ef755bd768ee9","verifierVersion":"0.1.0+18543036"}]}