Adding it to a group

Getting the bot into a group and flipping its switches is covered in Setting it up. This page is what it actually does once it's there.

Auto-scan: the short card

Someone pastes a contract address in the chat. The bot answers with a short card:

• header — token, chain, price, market cap

• the score

• both trade horizons with their estimated upside

• two buttons — 📊 TA Analysis and 👥 Holder Scan

That's it. Deliberately.

Every other scanner bot answers the same trigger with a full-size card, which is why busy groups end up as walls of bot output with a conversation occasionally visible between them. The short card fits three in a row without pushing the chat off screen — and it still shows entry and exit zones, which nobody else posts at all.

Tap 📊 TA Analysis and you get the whole card, chart included.

Tap 👥 Holder Scan and you get the holder and wallet-cluster report on its own, without opening the full card first — the distribution, the wallet quality and the funder-based cluster map. That is the question most groups actually have about a freshly pasted contract ("is the supply bundled?"), and it no longer costs a full TA card and a chart render to get to it. The short card stays where it is, so the other button still works afterwards.

The Holder Scan button only appears where the scan can actually run: Solana always, EVM chains when the cluster engine is enabled for them.

Three details that matter in practice

Old cards keep working. Auto-scan cards sit in the scrollback for hours or days, and the button on one of them still works long after it was posted. There's no session to expire and nothing to go stale.

Expanding an old card shows current numbers. Tapping a six-hour-old card runs a fresh analysis at the moment you tap it, rather than replaying the numbers from when it was posted. That's the honest thing for a card that may have been sitting there all night.

Repeats are deduped. The same token pasted again in the same chat within a short window (a minute, by default) doesn't get a second card. The check runs before a rate-limit slot is spent, so a repeated mention costs the group nothing.

Fresh launches are not auto-scanned

A token that deployed twenty minutes ago has no chart to analyse. Auto-scan passes it over in silence — no card, and no "skipped, too fresh" reply either, which would just be the same noise with different words.

There is a mode that answers those tokens as well, with a different card: no score and no entry/exit zones — those would be invented — but the on-chain read instead. It is reserved for the bot owner and cannot be switched on for a group. Two reasons, and the second is the one that decides it:

• That card is the thinnest thing the bot posts, and it is still being reworked.

• Every minutes-old contract pasted anywhere costs an upstream lookup, and every group draws on the same API budget. One busy chat pasting ten deploys an hour would spend what the rest of the groups need.

Two things stay unaffected:

• /don <address> always answers, fresh launch or not. Auto-scan governs what the bot says on its own, never what it does when it's asked.

• A token is still recorded as a call even when no card is posted, so /pnl and the leaderboard know who was first regardless.

And what looks fresh isn't always: an older token that has barely traded since it launched still gets its card, so a dormant contract waking up is never silently swallowed.

The other two group features

Feature	Command	Page
🃏 PNL cards	/pnl <contract>	PNL cards
🏆 Leaderboard	/lb	Leaderboard

Both are on by default and both are one admin tap from off.

They share a source: every /don in a group registers the first scan of a token in that chat, credited to whoever ran it. PNL cards read one row of that record; the leaderboard ranks all of it.

Rate limits in groups

Groups have their own windows on top of the per-user ones — 8 scans per minute and 120 per hour by default.

The per-minute one exists because Telegram itself throttles a bot at roughly 20 messages per minute per group, and one full analysis writes several times (status message, edits, chart). That ceiling bites first, whatever else is going on.

The per-hour one exists because nothing else caps a group: thirty active members each with their own hourly allowance could sustain hundreds of scans an hour in one chat and crowd every other group out. It deliberately leaves bursts alone — it bounds the sustained rate, which is the part that actually causes harm.

DMs are exempt from both, so anyone can always keep scanning privately.

When auto-scan stays quiet

In order of likelihood:

1. The bot isn't an admin. Telegram's privacy mode means a non-admin bot never receives ordinary group messages. Use the 🔄 Check admin rights button on the setup card, or /setup to reopen it.

2. Auto-scan is off. /setup, or /autoscan on.

3. The token is too fresh to analyse. Auto-scan skips tokens with no chart history; use /don, which always answers.

4. The group is rate limited. Wait it out, or scan in a DM.