Track record
How often the risk read is right
Every token someone checks on Trench Report is logged with the read it got at that moment, then watched for seven days. Here's what happened to each one, including every false alarm and every miss. Nothing on this page can be edited after the fact.
What counts as right and wrong
The same rules decide every token, automatically. They were fixed before any result came in, and a result can't be changed once it's recorded.
Every result
Newest first. Tap a token to open its current report.
Which warnings mattered
Check we didn't change it
Once each day ends (UTC), every read logged that day is fingerprinted with SHA-256 and the fingerprint is posted publicly on X. If anyone altered, added or back-dated a single read, the day's text would no longer produce the same fingerprint. The fingerprint covers the reads themselves; outcomes are decided later by the rules above, from public market data you can check on DexScreener.
shasum -a 256 day.txt on a Mac or Linux, or certutil -hashfile day.txt SHA256 on Windows, or paste the text into any SHA-256 tool. The result should match both the fingerprint here and the one posted on X that day.Changes to how reads are made
The outcome rules above never change. How the read itself is made can, and every change is listed here with its date. Reads logged before a change keep the read they got; nothing is re-scored after the fact.
- 5 October 2026: launch-risk checks added. Up to then, Solana reads were RugCheck's contract score alone, and this record showed the gap: all 23 Solana tokens checked within three hours of launch went bad, and 16 of them had been read Low signal risk because their contracts were clean. They were drained by the wallets that bought at launch. Reads now add points for how new the token is, insider wallets and holder concentration while it's under a day old, a deployer who keeps launching tokens, LP that can be pulled on EVM chains, and wallets behind earlier rugs on this record. On the 42 results resolved at the time, misses would have gone from 18 to 0 and correct High risk calls from 3 to 13, with no new false alarms; those numbers come from the same results the weights were set on, so the record from this date on is the real test.
- 5 October 2026: launch forensics for new Solana tokens. For tokens under a day old, reads now also check, from Solana's own transaction history (via Helius), how much of the supply wallets other than the deployer bought in the launch block, and whether today's top holders got their first SOL from the same wallet. The wallets funding such clusters are remembered, so a later launch funded the same way is linked to this record's earlier rugs. These checks have no history on this record yet; they're scored modestly and will be reviewed against outcomes.
What this does and doesn't show
- Only tokens people chose to check. Not every token launched, and not a random sample. People tend to check tokens they're unsure about, which can push every number up.
- The first read only. A token is logged the first time anyone checks it here. A read that changes later (a mint authority revoked, say) doesn't change the logged one.
- Most memecoins fail whatever the read says. That's why the base rate is shown beside every number. A High risk read is only useful if High risk tokens go bad more often than tokens in general.
- Low signal risk is not "safe". It means the structural checks found nothing. A token can pass every check and still be sold into or abandoned.
- Not financial advice. This page measures a tool. It doesn't tell you what to buy or sell.