Skip to content

Microchain: a tap you cannot open wider

An instrument has to exist before it is needed. A request made in a crisis is cold, binary, and about loyalty, so it fails. A position opened in peacetime — with a price, a history, and a visible right attached — is a transaction, and a transaction survives the moment it gets used.

Microchain is that instrument, in one Telegram channel and one pair of TON contracts. A bonding curve prices the token continuously. Holders vote with locked tokens. One pinned message carries a signed snapshot anyone can verify offline.

And the part worth your attention: the issuer’s own claim on the money is a rate, not a promise.

Buy in. The curve quotes you a price and mints against the TON reserve that buyers themselves form. The price is a position on a curve, not a valuation — each token costs more than the last, and the arithmetic is public.

Read the ceiling. One command tells you the most the issuer could put into the curve today, and how long the rest takes to arrive. It is a number off a clock, not a character reference.

Exit whenever. Burning returns TON from the reserve at the curve price. No lockup, no queue, no permission.

Vote with weight. Lock tokens on an option and your weight is the square root of the stake — four times the tokens buys twice the weight. On an ordinary poll the stake unlocks at the close; on a tap poll it is held a further day, so the capital that decided the outcome stays under it.

Propose a cut. Any holder can put lower the tap or stop the tap on the ballot by bonding tokens. The issuer is not in that path and cannot veto the result. The bond comes back if the poll reaches quorum and stays with the contract if it does not, which is what stops the ballot filling with noise.

Carry out the result yourself. When a tap poll closes and wins, the message that executes it checks no sender at all. Your wallet can be the one that moves it. So can a stranger’s.

In the Telegram bot:

/link EQ... point the bot at your wallet — do this first
/price the curve, the reserve, what backs a token
/tap the rate, what is released, what is left
/buy 5 a link that buys 5 TON worth
/sell 10 a link that exits 10 tokens
/balance what you hold and what it exits for
/polls what is open right now
/vote 1 0 25 lock 25 tokens on option 0 of poll 1
/propose cut | 72 | 40 put "cut the tap to 40 tok/day" to a 72h vote
/propose stop | 168 | 5 put "stop it for good" up, bonding 5 tokens
/apply 1 carry out poll 1 once it has closed and won
/withdraw 1 unlock your stake after poll 1 closes
/verify the addresses, the signing key, the rights hash

Every one of those hands back a link you sign. The bot never holds your keys and never moves your tokens.

In the Mini App at microchain.kaido.team: buy, sell and vote on the main surface, apply under past votes once a tap poll has closed and won, and the curve drawn with the ceiling shaded onto it — the solid block is what the tap has already released, the dashed outline is where it can still reach. Proposing is the bot’s job.

Deployed on mainnet, and empty — nothing has been bought yet, so the curve still sits at its opening price and the first buyer gets it:

master EQAuQZv1EY3Y6dBsAJ7yVHa85ILRXpNYpzk_c2-eCdnuZAC_
governance EQAsy6pG990MPt5bfMF1Yt-b_kXZ2TmtWeuk3Q0xVeuy9yg7
tap 1000 tokens over 12 months — about 2.74 a day

The tap has been running since deploy regardless, so the issuer’s claim accrues whether or not anyone buys — which is the one number to check before you do.

The obvious way to limit a founder allocation is to cap how fast it can be sold. That does not work, and the reason is worth knowing before you trust any scheme that claims it does.

A token that already exists does not remember who received it. The only check available is is this seller the issuer? — and one transfer to a fresh wallet answers no. A limit that costs one extra transaction is not a limit.

So the allocation is never created up front. There is a ceiling and a rate, and nothing exists until it is claimed. What can be taken is what the clock has released:

claimed ≤ elapsed × rate

That holds whatever wallet anyone uses, because the constraint sits on creation rather than on identity. Issuance is the one chokepoint a jetton master genuinely owns.

Here is the part that generalizes past this project.

Votes are cheap. Buy tokens, vote, wait for the close, sell back — the curve returns what it took, so you paid gas and the cost of capital sitting still for a week. You did not buy a vote. You rented one.

Which means the question is not how to make voting harder. It is what you put in the set of rights.

Hand a vote a power that is profitable to somebody, and you have listed that power for sale at the price of renting money.

So holders can only narrow the tap: lower the rate, or stop it permanently. Raising it is available to no one — not the issuer, not a unanimous poll. Nothing is left in the effect space that anyone would want to buy, so the attack stops making sense: not because it got expensive, but because there is nothing to win.

That is also the only reason the issuer could give up the veto over money at all.

A vote somebody has to execute is not a power. It is a memo.

Someone holds the credentials. Whoever holds credentials can refuse, and no amount of arithmetic changes that — a contract cannot change a picture in Telegram, and promising otherwise is a lie.

But the tap lives where the votes live.

So the message that carries out a tap decision checks no sender. Everything it asserts is visible on-chain: the poll closed, quorum was met, no veto, yes won. A holder can send it. A stranger can send it. Someone who dislikes the outcome can send it, and it still lands.

No key. No permission. That is a power.

A price that moves on real money. A vote whose weight is the square root of what you put behind it. An exit that needs nobody’s approval. And a ceiling on the issuer’s side of the table that you can read off a clock and tighten by agreement.

A promise cannot be checked.

A rate can.