Getting started
Use Marden to discover Solana launches, inspect token evidence and keep a watchlist. Start with the public terminal; run a token check when you need the full report.
How to use it
- Open the terminal to see launches and tokens with the numbers attached to them.
- Open any contract for a summary. A summary is free and needs no sign-in.
- Run a check to open the full report: exit cost at several sizes, pool depth, developer history, and what was recorded when the token was called.
What it will not tell you
It will not tell you a token is safe, and it will not predict a price. It reports what was measured, when, and from where — including the cases where nothing could be measured.
Selection
Two different decisions, kept apart.
| Stage | What it means | What it is not |
|---|---|---|
| Eligible | A market reading at or above the floor puts the token in the review queue. | Not a call, and not a claim about quality. |
| Called | Every selection gate was met and the decision was written down with the reading it was made on. | Not a promise that the token goes up. |
| Legacy record | Written by an earlier policy that only knew the floor. | Excluded from Calls rankings and performance statistics. |
The gates
A call needs bundle and cluster share at launch, developer history on the deploying wallet, pool depth against a $500 exit, and the number of distinct tracked buyers. Verified selection is not active yet. Bundle evidence and the selection rules must be available before a token can qualify. Discovery listings are not endorsements.
The baseline
The reading the decision was made on becomes the baseline, and it is never rewritten. A token that triples afterwards cannot move its own starting point, which is the only way the record can be read later.
Thesis & performance
Three kinds of information, deliberately separate.
| Field | Where it comes from | How to read it |
|---|---|---|
| Thesis | An interpretation written from the measurements taken when the call was made: round trip, exit depth, pool count, developer history, early-buyer count, tracked buyers. | An argument about why that evidence was enough to call the token — not a claim about why the price moved. |
| Project narrative | The description the project published in its on-chain metadata, quoted and attributed. | Their words. Many projects publish nothing, and those rows say so. |
| Performance | Market capitalisation now and at its highest since the call, divided by the baseline. | A market-cap multiple, not the profit an exit would have paid. |
Tokens that fell under the floor stay in the list. A missing reading is shown as missing; it is never counted as zero, and never silently removed.
What one record looks like
| Field | Example | Reading it |
|---|---|---|
| Baseline | $4,200 | The market cap at the moment of the decision. Fixed for the life of the record. |
| Now | $12,600 · 3.00× | The latest observation, divided by the baseline. Moves every time the token is measured. |
| Peak | $31,080 · 7.40× | The highest observation since the call. It stays at its highest even if the token falls back. |
| State | left the table | Out of the discovery window but still being measured, so the record does not freeze. |
| Missing data | 1 missing | Which readings are absent. Hover the cell for the list; nothing is filled in. |
Terminal & reports
Terminal
- Launches — tokens captured from the curve, with progress and the SOL sitting in it.
- Tokens — tokens with a pool, with the exit cost and pool depth attached.
- Pools — where liquidity sits and how likely the price stays inside a range.
Filters are open by default. Market filters take your own numbers, and k or m after a number means thousand or million. Filtering and refreshing never spend a check.
Report
Open a token to view its summary and detected pools. Select Run token check to open available exit estimates, liquidity analysis and developer history. Missing measurements and unverified pool ownership are labelled; not every token has a complete report. Reloading an opened report during the same access window is free.
Watchlist
Contracts you keep coming back to. Signing in with a wallet is only needed to keep the list on the server, so it follows the wallet to another browser. Browsing does not require signing in.
- Up to 50 contracts per list.
- Without a sign-in the list lives in this browser and disappears when it is cleared; the page says which of the two you are looking at.
- One wallet sees its own list. A different wallet sees a different list.
Access & limits
Free: 5 successful token checks per 5 hours per IP address. The window opens at the first check that succeeds, and the report shows the time it resets.
What counts: opening a contract with a check, including deliberately checking the same contract again. What does not: reloading a report already opened in this window, a failed check, a retry or double click with the same request, and the five-second table refresh.
Signing in saves your watchlist across devices. It does not increase or reset the free check allowance. People using the same public IP address may share an allowance.
Holder: 350,000 tokens in a wallet we can read would raise the limit to 50 checks per 5 hours per wallet, drawn from its own allowance. Not active yet — the token has not launched, so holder access cannot be activated or purchased yet. Free access remains available.
Sources, refresh & limitations
Where the numbers come from
- Market capitalisation and pool data from public market feeds, each row carrying the time it was read.
- Curve state and supply read directly from the chain.
- Project descriptions read from the mint's own metadata, quoted as written.
Refresh
The terminal, homepage and watchlist request updates every five seconds. Calls checks every 30 seconds. Updates pause in hidden tabs and slow down after errors. Page refresh frequency does not determine source freshness: market feeds, on-chain readings and historical models update on different schedules. Check the displayed source time before using a figure.
Known limits
- Launches whose supply could not be read are listed separately as supply unread and held out of token analysis: no token row, no call, no check. An unread supply is not a normal launch and is never treated as zero.
- Launches whose curve account could not be read are held out of the launch table until it can be read, because there are no curve numbers to show. They are not mayhem launches: only the flag on the curve account marks a mayhem launch, and both counts are published at /api/health rather than dropped without a word.
- Not every project publishes a description, and none is written on their behalf.
- Third-party intelligence providers are not integrated. Nothing here claims a provider verified a token.
- Market capitalisation is an estimate built from price and supply; it is not the cash you can withdraw.
Endpoints and their gates are documented separately in the API reference.