Innovative payment smart contract with reputation system run via telegram bot

Wallet address to wallet address crypto transfers would be orders of magnitude better than fiat payments if a reputation system were built in. Outside of blockchain it is impossible to embed reputation in payments. This blockchain advantage has never been leveraged. Until now. Until Free TON.

User stories:
This above system can be achieved with a smart contract and telegram bot - lets call it “repbot” - as follows:

  • Bob and Alice have joined repbot
  • Using repbot Bob sends Alice 10 crystals by sending the bot her telegram handle and then the amount he wishes to send;
  • Alice receives a notification from repbot about the payment but she cannot take the payment to her wallet until a rating is provided by Bob. Neither can Bob provide a rating until Alice approves the payment. The notification she received asks her to approve the payment. She approves it. Once approved by Alice, Bob cannot now retrieve his crystals from the smart contract.
  • Repbot notifies Bob that Alice has approved the payment but that he must rate the transaction, for example, 1-5. Maybe Alice is a barista and this whole exchange happened in her coffee shop, or perhaps Bob was buying a car (a very small toy one!) and the exchange is happening in person. Distance transactions are less trivial and are discussed below. Bob is satisfied that all is in order and gives Alice a 5 rating.
  • Once rated, Alice needs to rate Bob in order to access the 10 crystals. She rates bob 5 in response to the prompt from repbot. The crystals are automatically sent to her wallet.


  • The purchase of crypto has been banned. Based on their respective reputation scores Alice and Bob meet and exchange cash for TON crystals.
  • An online market place is built by the TON community. Reputation scores and telegram addresses of sellers are displayed. 50% of the purchases are conducted face to face and the reputation scores ensure a level of safety that buyers did not before have. Where the items need to be shipped buyer and seller decide by discussing the transaction on telegram how they want to utilise repbot. Usually, if the buyer has a high reputation, the seller is content to accept the payment and ship in the knowledge that the buyer cannot access the paid funds.

Attack vectors
To prevent people gaming the reputation system in order to abuse it, reputation given could be weighted depending on the reputation of the account giving it. For example, a reputation score given by an account with established reputation achieved through multiple transactions would increase the recipient’s reputation score more than that received from a new account. Perhaps a fee could be levied on the transaction - instead of rent seeking platform, a strong reputation system is incentivised.

Contest Scope
Task 1: improve the overall design including highlighting new issues and attack vectors. 1000 TON divided across 5 overall prizes
Task 2: develop the specs for smart contracts such that a developer could construct it. 2000 TON divided across 3 overall prizes
Task 3: develop the specs for the telegram bot such that a developer could construct it. 2000 TON divided across 3 overall prizes

If a reputation system was embedded in a blockchain based payment solution, a step change over what e-commerce, banking etc. provides would be achieved. This proposal moves away from centralised transaction systems completely and provides greater hope for decentralisation and an important onramp for TON crystal adoption.

Contest status: this contest has not yet been posted. I am seeking community feedback beforehand. Plus tbh I don’t know what a “proposal key” is. Maybe someone can help with that too.


Great idea! If implemented correctly, it will make a splash

it looks like I am using Alibaba to purchase a product or using localcrypto or localbicoin to buy crypto.
Am I right?

Hi @wealth20xx. In Europe, competitors to eBay emerged which trimmed all the “big tech” complexity away and simply allowed people to post, share contact details and work out the details of their transaction between themselves. One example is The main use case is person to person transacting e.g. buying a secondhand boat. They are a bit like craigslist, I guess (I lived in New York 5 years) but more managed/moderated. With this proposed contest I posit that the Free TON blockchain could level-up such commerce by embedding reputation scores in transactions. I go so far as to argue that this is the next leap in online transactions innovation. Furthermore, it is an innovation that centralised incumbents cannot technically achieve meaning it is an opportunity for crypto to add value that non-crypto providers can never add. Alibaba is a rent-seeking e-commerce platform. The proposed technology is marketplace agnostic - any developer, in any area, or for any niche could leverage the ability to exchange crystals easily via telegram, and the robust user reputation (simply read with getters from the blockchain) to create a market. Localbitcoin is pretty much suppressed these days in its capacity to offer cash-bitcoin trading. Again, the proposed technology allows anyone to launch a “local-free-ton” site connecting buyers and sellers in the context of reputation and a novel transaction process. Another reason why I think this concept is a great fit for Free TON is the availability of rewards for “most innovative marketplace that does x, y, or z” and that utilise this solution - such contests would spur revenue model innovation for such marketplaces.

That idea isn’t new of course :slight_smile:
But why not? The only one problem - we have no market so far :sweat_smile:
And think about to remove telegram bot from your scheme. Reputation must be stored on-chain…

Thank you @aicracy. The bot would be the means with which users interact with the Free TON blockchain - reputation scores recorded on-chain. Can you elaborate on why this should not be so? Also, I never saw this idea described or implemented before. Could you point me somewhere where I can see other versions. I was quite inspired with Button Wallet’s implementation some time ago, and wished through reputation scoring such an approach could surpass the utility that “normal” transactions afford.

Not really, but that is logical in general. In real life that function is a part of notarial act, as so-called notarial deposit. You suggest to replace the notary with a smart-contract. That’s all.

How Bob is protected against to be rated with “1”? What forces Alice to rate with “5”?

Alice can rate Bob 1. And you or I can rate a good Uber driver 1 too if we want to. But Bob probably wont buy from Alice again if that happened and it was not justified. In addition, the weighting of ratings made by users who themselves have few ratings is low so the damage done to Bob is lessened. It might also be possible to make poor ratings more costly for users - all could be explored.

If I travel by train and buy pies from old woman at the station, I’m sure that I will never buy anything here again :slight_smile: Just one scenario…
I can also repeatedly compromise addresses one after another, as a buyer, especially if the product or service does not require physical delivery to my address and I can remain anonymous.

So I have the feeling I’m already taking part in proposed contest :rofl:

Its great. But remember that the opposite is true. Existing centralized reputation systems have made unassuming enterprises such as this “old woman” into success stories. Your argument suggests that the outlier (the person wanting to destroy the old woman out of spite) outnumbers the authentic - those that give good feedback for good service.

How? The product or service providers rating of you will suppress your ability to give malicious negative feedback over time.

For every transaction I use new wallet address with blank reputation.

Identity and reputation are not the same. Someone with no identity can still have reputation. Its just tied to their telegram handle in this case.

I agree. And as seller you’re interested in maintaining of the reputation of your wallet address. But not as a buyer and last hand in that story.

And how will events develop in case of a return of goods?

Does this effectively recreate the “bug”:

  1. A product or service can be delivered digitally (not physically). Lets take a silly example and say it a photo of my cat.
  2. Someone despises me and my cat. Hmm maybe a competitor in the cat photo market.
  3. They pay for my cat photo and then give me a bad review.
  4. They create a new wallet, pay for my cat photo and then give me a bad review.

There might be cases where the economics of this stack up. But the weighting of feedback given is key. For example, the design of this system might be that 20 feedbacks from new wallets carry the same weight as 1 feedback from a wallet with a high reputations score. So the economics of this attack just got more difficult.

As they say, decentralization is hard.

Is it not fair to say that empirically buyers generally leave fair feedback? This is what we see right? I think we need to try to fix attack vectors as opposed to argue that this statement is incorrect.

You rely on the honesty of people, in fact on the human factor, while trying to build a trustless system? :slight_smile: I see a little paradox in it.

It really does not claim to be a replacement for escrow. Neither by design nor intension.