Onchain analysis becomes easy with
@Hertzflow_xyz , It's building a forensic layer between the CA and the decision.
Analyzing Hertzflow skills:
skill.hertzflow.xyz
Most token analysis starts with price, volume, market cap and liquidity.
HertzFlow Skills takes a different approach: start with the wallets and follow the supply.
The Alpha Forensic skill is built for Binance Alpha-listed tokens and can investigate BSC, Ethereum, Arbitrum, Base, Polygon and Optimism.
You paste a CA and the pipeline can trace project-linked wallets, pre-launch distribution, downstream transfers, recent 72h anomalies, wallet clusters, funding sources and confirmed insider sell-out activity.
The useful part is the classification of the current holder structure.
Instead of treating every wallet as “retail,” the report separates roles such as deployer, treasury/vesting, quiet insider, partial dumper, CEX custody, DEX liquidity and other holders.
It also traces the project's wallet lineage showing which wallets received tokens from the deployer, how much they still hold and how much has already moved.
Then it looks at real distribution, using confirmed insider transfers into CEX deposit addresses and direct DEX sells rather than simply assuming that every large transfer equals selling.
There is also a 3-way chip view:
📌Operator / CEX pool / verifiable retail.
That gives you a better picture of where the circulating supply actually sits.
Another useful layer is the anomaly analysis.
The system groups meaningful movements into pre-launch OTC seeding, downstream redistribution and recent 72-hour activity, making it easier to identify whether distribution appears finished, ongoing or recently starting.
HertzFlow also detects connected wallet clusters and provides funding-source attribution for high-value wallets, helping distinguish wallets that received tokens through minting or P2P transfers from wallets that actually acquired them through DEX activity.
Then comes the practical side:
The report calculates a maximum single-tranche entry size at 5% slippage using Binance Alpha's official depth data and generates a monitoring list of wallets that can be tracked afterward.
And importantly, the architecture isn't simply ask an AI what it thinks.
The deterministic Python pipeline owns the underlying data and evidence.
The LLM fills predefined narrative slots.
A validator checks those outputs against the evidence before the final report is rendered.
So the product isn't trying to replace the trader with an AI buy/sell signal.
📌📌📌📌📌📌
Don't just ask where price is going.
First understand who controls the supply, where the tokens are moving, and what the on-chain evidence actually says.