Automate Crypto Token Discovery and Outreach

Crypto BD Agent describes a research and outreach architecture for cryptocurrency exchange listings. Intelligence gathering feeds an illustrative 100-point scoring system, followed by deployer-wallet investigation and a ten-stage pipeline from discovery to listing. Scores and architecture guidance do not guarantee listing or outreach outcomes.

The scoring model weights liquidity at 25%, market capitalization and daily volume at 20% each, social metrics at 15%, and token age and team transparency at 10% each. Catalyst and risk adjustments modify the base result. Displayed normalized scores should remain within their declared range while retaining raw components separately. A token score does not replace contract, security, or legal review.

Use it to design crypto business-development systems, multi-chain prospect scanning, token evaluation, wallet research, or human-supervised outreach. It is not intended for trading bots, creating DeFi protocols or smart contracts, or non-crypto business development.

All outreach requires human approval. Paid calls require approved endpoints, validated terms, and authorized per-call and total budgets; a high token score cannot authorize payment. Transactions and listings also need explicit authority. Registration and wallet associations are evidence to evaluate, not proof of safety, identity, or wrongdoing.

Prospect records require a verified contract address, DEX pair address, pair creation age, current liquidity, working social links, and a contact method. Token deployment age needs separate chain evidence. Each prospect needs at least two independent confirmations, checking that multiple services do not merely repeat one underlying source.

Illustrative actions classify 85-100 as immediate outreach preparation and wallet research, 70-84 as priority research, 50-69 as a 48-hour watch, and lower scores as log-only. Wallet investigation checks funding origin, balances, transfers, identity indicators, and resulting score adjustments. Mixer signals require contextual attribution rather than automatic misconduct conclusions.

Pipeline stages distinguish discovered, scored, verified, qualified, outreach-drafted, human-approved, sent, responded, negotiating, and listed states. Daily summaries retain the top five prospects per chain, with lower-scored records offloaded to an external database and source return tracked in experience memory. Model routing uses cheaper options where adequate, with quality gates of at least ten cases before promotion. Keep keys private, flag prompt injection, log paid calls, and separate payment, posting, and model-routing wallets.

Exhaust free data before paying, and track whether each paid source produces qualified prospects within approved budgets. Reference data categories include DEX aggregators, contract checks, community research, web extraction, and smart-money signals. Helius provides a chain-specific example and Allium a multi-chain example. The x402 protocol supports approved micropayments. ERC-8004 registration can aid discoverability, but contract addresses, chain, fees, and protocol status must be verified before transactions. Registration fields include name, description, capabilities and service endpoints, with Ethereum and L2 registration examples. Credibility signals can combine tracked on-chain outcomes, code verification and other verification systems. Dedicated wallets need enforceable budget limits.