free import audit
Did your tax software get Hyperliquid right?
Upload the transaction export your tool produced — Koinly, CoinTracker, anything with a CSV — and we recompute the same window from raw venue data. Then we line the two up, stream by stream. The file is parsed in your browser; only counts are sent.
In Koinly: Wallets → Hyperliquid → download the transaction list as CSV. In CoinTracker: Wallets → Hyperliquid → export history. Only row counts and dates are sent — the file itself stays on your machine. (synthetic Koinly file audited against a public address — the recomputation itself is real).
what gets caught
Koinly documents a 10k-transaction sync cap per wallet. An active perp account passes it in weeks — the audit flags a file sitting at exactly that count.
Funding settles hourly on its own endpoint — 8,760 events per year per open position. A fills-only import shows $0 where the venue shows thousands.
The venue's fill API retains ~30 hours. Anything imported through it is a few days pretending to be a year — the date row in the verdict shows it directly.
- — Counts compare one venue event to one export row. Aggregation lowers counts legitimately; a missing stream is the signal.
- — Evidence, not verdicts: we print what the venue shows and what your file contains. The classification decision stays with you and your preparer.
- — Found a real gap? The paid report produces the corrected record with basis, or see why imports break for the mechanics.
questions filers actually ask
How do I export my transactions from Koinly?
In Koinly: open Wallets, pick your Hyperliquid wallet, and download the transaction list as CSV. Upload that file here — not the tax report PDF. Koinly documents a 10,000-transaction sync cap per wallet; if your file sits near that many rows, the overflow simply didn't make it in.
How do I export from CoinTracker?
In CoinTracker: open your Hyperliquid wallet and export its transaction history as CSV. The same upload works — we detect the format from the header row automatically.
Is it safe to upload my tax CSV?
The file never leaves your browser. The audit parses it locally and sends only aggregates — row counts, date range, income totals — alongside your public address. Your address is public on-chain data; the CSV contents stay on your machine.
What does the audit actually compare?
We recompute your account's activity from raw venue and chain-indexer data for the same window — every fill, every hourly funding settlement, the ledger flows — then line the counts up against what your file contains. Coverage gaps (rows vs events), the funding stream (income rows vs funding events), and history depth (your file's first date vs the window start). Where a stream is absent entirely, that absence is printed.
My import shows fewer rows — does that prove it's wrong?
Not by itself — some tools legitimately aggregate many events into one row. The meaningful signals are a whole missing stream (zero income rows on an account with thousands of funding events), a file that stops exactly at a documented sync cap, or a date range that ends months before yours did. The audit prints the evidence; the conclusion is yours.