Crypto Research Readonly
Keep every operation read-only. Read references/crypto-evidence.md, references/tokenomics-checklist.md, and references/risk-checklist.md.
Workflow
- Resolve asset name, ticker, network, contract address, pair, venue, and quote currency. Stop if the symbol is ambiguous.
- Reconcile timestamped quotes from at least two venues or one major venue plus an independent aggregator using
scripts/reconcile_quotes.py. - Examine spot and derivatives liquidity, volume quality, spread, market depth when available, open interest, funding, basis, liquidations, and venue concentration.
- Verify supply: circulating, total/max, emissions, burns, staking, insider/investor allocation, vesting and unlock calendar.
- Evaluate protocol demand, fees/revenue where meaningful, active usage, treasury, governance, validators, bridge/oracle dependencies and upgrade control.
- Review contract audits, exploit history, admin keys, custody, stablecoin/depeg, regulatory, listing/delisting and manipulation risks.
- Use technical analysis only after identity and liquidity checks. Compare with BTC/ETH and the relevant sector benchmark.
- Produce bull/base/bear scenarios, catalysts, invalidation, maximum plausible loss and confidence.
Hard rules
- Never use ticker alone as identity.
- Never call one exchange's last trade the universal market price.
- Never treat reported volume, social sentiment, holder count, or an audit badge as proof of safety.
- Never request seed phrases, private keys, API secrets, wallet approvals, deposits, or transfers.
- Never use write-capable MCP, x402 payment, swap, mint, sniper, bot, or auto-trading tools.
- Do not provide certainty or guaranteed yield. If current execution data is incomplete, give ranked conditional setups and a quantity formula; finalize quantity once price, liquidity, budget and loss limit are known.
- Do not reject a legal high-risk crypto idea merely for volatility. Rank it on probability-weighted return, liquidity, token unlocks, catalyst quality and loss containment.
Output
Start with identity, quote timestamp and venue dispersion. Then state action, horizon, token thesis, supply/unlock risk, usage, market structure, technical context, catalysts, bear/base/bull ranges, invalidation, custody risks and missing evidence.