You paste a mint, wait a few seconds, get a red/yellow/green summary. Full source lives here (copy and run): solcheck.py on Telegraph.

Technical walkthrough of how the checker is built. Not a buy bot. Not financial advice. Plenty of tokens still die after a clean-looking printout.

Why bother with a CLI

Browser tabs are fine once. A script is better when you re-check the same mint five minutes later, or want JSON for a sheet. With one command you can:

• re-check the same mint after five minutes
• pipe --json into a spreadsheet or a daily snapshot
• keep the logic versioned next to a personal checklist

python3 solcheck.py SoMeMiNtAdDrEsS...
python3 solcheck.py SoMeMiNtAdDrEsS... --json

Python 3 only. urllib + json. Zero pip dependencies, zero API keys, zero accounts.

Architecture in one paragraph

For each mint the script fans out five HTTP requests, merges the answers, computes a handful of derived fields (top-10 holders with pools stripped out, LP lock %, age, buy/sell skew), then scores each check into red / yellow / green. Failed sources become warnings and n/a fields; the process does not abort.

mint
  ├─ RugCheck report ── holders, risks, LP, creator, score
  ├─ DexScreener tokens ── best-liquidity pair stats
  ├─ Jupiter price v3 ── fallback liquidity / createdAt
  ├─ Jupiter token search ── name/symbol/holderCount fallback
  └─ Solana RPC getAccountInfo ── mintAuthority, freezeAuthority
           │
           ▼
     merge → flags → text or JSON

Request layer

Rate limits happen. The helper retries 429 and 5xx with linear backoff and caps the error string so a stack trace never floods the terminal:

def fetch(url, body=None, retries=3, timeout=20):
    data = json.dumps(body).encode() if body is not None else None
    hdrs = dict(UA, **({"Content-Type": "application/json"} if data else {}))
    err = None
    for i in range(retries):
        try:
            req = urllib.request.Request(url, data=data, headers=hdrs)
            with urllib.request.urlopen(req, timeout=timeout) as r:
                return json.loads(r.read().decode()), None
        except urllib.error.HTTPError as e:
            err = f"HTTP {e.code}"
            if e.code in (429, 500, 502, 503, 504):
                time.sleep(1.5 * (i + 1))
                continue
            return None, err
        except Exception as e:
            err = type(e).__name__ + ": " + str(e)[:80]
            time.sleep(1)
    return None, err

DexScreener can return several pairs for one mint. Keep the one with the highest USD liquidity:

pair = max(
    ds,
    key=lambda p: (p.get("liquidity") or {}).get("usd") or 0,
) if isinstance(ds, list) and ds else {}

Derived fields that matter

1. Authorities (RPC wins)

info = (((rpc or {}).get("result") or {}).get("value") or {}).get("data", {})
info = info.get("parsed", {}).get("info", {}) if isinstance(info, dict) else {}
mint_auth = info.get("mintAuthority") if info else rc.get("mintAuthority")
freeze_auth = info.get("freezeAuthority") if info else rc.get("freezeAuthority")

Active mint authority = someone can print more supply. Active freeze authority = someone can freeze your token account. Both are hard reds when present.

2. Top-10 holders without fake whales

RugCheck's knownAccounts tags AMM and LOCKER addresses. Markets expose pool pubkeys. Drop those, then sum the first ten remaining percentages:

excl = {a for a, v in (rc.get("knownAccounts") or {}).items()
        if v.get("type") in ("AMM", "LOCKER")}
# also add market / LP account pubkeys from rc["markets"] ...
real = [h for h in (rc.get("topHolders") or [])
        if h.get("owner") not in excl and h.get("address") not in excl]
top10 = sum(h.get("pct") or 0 for h in real[:10]) if rc.get("topHolders") else None

Thresholds used by the flags: ≤15% green, 15–30% yellow, >30% red. On huge old coins, CEX and treasury wallets can inflate this, so treat a red on BONK-class names with context.

3. LP lock and bonding curve

While a token is still on the pump.fun curve, "LP locked" is the wrong question. The script marks that as yellow (still on pump.fun bonding curve) and only applies the lock % after migration. Lock ≥90% green, 50–90% yellow, under 50% red.

4. Dev and RugCheck extras

• creator balance as % of supply (from RugCheck + RPC decimals)
• count of other tokens the same creator launched (>3 → yellow "serial launcher?")
• graphInsidersDetected and insider-flagged top holders → red
• each RugCheck risk name, mapped by level (danger → red, else yellow)
• age under one hour → yellow
• liquidity under $10k after migration → yellow
• buys/sells > 8 → yellow (possible wash)

Sample outputs from a live batch

On 2026-11-05 at about 20:11 Europe/Moscow I pointed the same code at 28 Solana mints from DexScreener boosts/latest profiles. Result that hour: 19 RED, 5 YELLOW, 4 GREEN.

Clean-looking basics: $DESSA (npz2FhoufQ7pKqhGpaipYWacrAkidyW2yDpVWSvpump):

• age 10.2 days, mcap ~$33.7M, liq ~$546k, holders 37.6k
• mint/freeze revoked, top-10 ex pools 2.7%, LP locked 100%, graph insiders 0
• verdict GREEN (still: check bundles on trench.bot / Bubblemaps yourself)

Caution: $GHOST (42qKoGmEPTRJsa5jWrYwpydL8RkBwMvg2ykYVoerpump):

• age 23.1 h, mcap ~$210k, liq ~$38k
• top-10 ex pools 19.5% (yellow) + RugCheck "High holder correlation" (yellow)
• verdict YELLOW

Concentrated + insider graph: $Raul (C1dNqHWwrLC6Hmtbr75BL3FmH4UCN3ztam7VDtwhpump):

• authorities revoked, LP 100%, but top-10 ex pools 38% and 6 graph insiders
• verdict RED

Authorities and LP alone would have looked fine on $Raul. The holder and insider steps catch what the chart does not.

Gaps on purpose

Bundle and sniper detection has no solid free API I trust. The script refuses to invent one. It prints:

https://trench.bot/bundles/<MINT>
https://app.bubblemaps.io/sol/token/<MINT>

Social copycats, fake sites, and Telegram drama also stay outside the CLI. Pair the script with a short manual checklist and you are faster than either alone.

Try it

Grab the file from the Telegraph script page, then:

python3 solcheck.py <MINT>

Re-run before you size a trade. Memecoin stats rot quickly.

Disclaimer

For education. Not financial advice. No profit claims. Trading memecoins can wipe the wallet.

Disclosure: Axiom referral link. New signups through it get Axiom's documented 10% trading-fee discount; I earn a share of fees. Open before you register: https://axiom.trade/@anteikq