PonsV2LauncherToken $AP

Risks (11)

• HighClaim contradicted: Platform tokens can be burned to vote on an agent's preset, permanently removing them from supply with a minimum of 1,000 platform tokens per vote.claims-mismatchDeterministic

  The project claims token burns / buybacks, but no tokens of this contract were ever sent to a burn address or destroyed.

  Evidence

  • Xhttps://x.com/agentic_pad/status/2117716859587883343
    Show source

    Every 4 hours anyone can vote on an agent's next preset by burning at least 1,000 platform tokens. The most burned wins the next window. Every vote takes tokens out of supply for good.

  • on-chainTransfer events of 0xb5ab7c318797698362cee3307b58c16be271a59a to burn addresses + balanceOf(0x…dEaD)
  • Xhttps://x.com/agentic_pad/status/2117702515248488698
    Show source

    What the agent does with its money: • buys back its own token and burns it

  Confirmed by: facts

• HighClaim contradicted: 1% of every trade funds the token's AI agent to buy back, airdrop to holders, and trade.claims-mismatchDeterministic

  The project claims token burns / buybacks, but no tokens of this contract were ever sent to a burn address or destroyed.

  Evidence

  • webhttps://agenticpad.family/
    Show source

    1% of every trade funds the agent; it buys back, airdrops to holders and trades on its own.

  • on-chainTransfer events of 0xb5ab7c318797698362cee3307b58c16be271a59a to burn addresses + balanceOf(0x…dEaD)

  Confirmed by: facts

• HighCreator-declared wallets bought 27% of supply at launchinsidersDeterministic

  4 wallet(s) linked to the team (declared by the creator at launch, exempt from the snipe tax) bought 27% of the supply in the first blocks — an insider allocation that real buyers did not get. They now hold 1.24%, so current dump pressure from them is low; most of it has already been sold.

  Evidence

  • on-chaininsiders: 4 wallets, fundedBy declared by creator at launch (snipe-tax exempt on curve; creator 0x54fca90e647aed824fc47d0b968dbf132e7cdd24), bought at launch 27%, pctSupplyHeldNow 1.2434
  • on-chaintoken.insiders[0]: wallets 0x54fca90e647aed824fc47d0b968dbf132e7cdd24, 0xc2df3da3be20084ccd2350a9608243124bb1dce1, 0xde49332a744b027af76b72ede6415f3880395cd3, 0x04453fba30babcb009a8cb821d5593b7a70839f0; pctSupplyHeldNow 1.2434
    Show source

    declared by creator at launch (snipe-tax exempt on curve; creator 0x54fca90e647aed824fc47d0b968dbf132e7cdd24)

  • on-chaintoken.bundles: block 105273923 wallets 0xde49332a744b027af76b72ede6415f3880395cd3, 0x04453fba30babcb009a8cb821d5593b7a70839f0 pctSupply 26.9999
  • on-chaincurve deployer (creator fee recipient) = 0xc2Df3dA3Be20084CCD2350A9608243124BB1dce1

  Confirmed by: facts, claude-opus-5.5, grok-4.7

• HighLaunch sniped: 46 wallets took 94.4% of supply in 10 secondssnipersDeterministic

  The launch was captured: 46 wallets bought 94.4% of the supply within 10 seconds. A small, fast group set the early price and could sell into everyone who bought later — bad for launch fairness and a strong insider/bot signal. They now hold 0% of supply, so current dump pressure from them is low — most of it was already sold, i.e. they exited on later buyers.

  Evidence

  • on-chainsnipers: windowBlocks=100, wallets=46, pctSupplyBought=94.4014, pctSupplyHeldNow=0
  • on-chaintoken.snipers: windowBlocks 100, pctSupplyBought 94.4014, pctSupplyHeldNow 0
  • on-chaintoken.clusters: funder 0xcdd37ada79f589c15bd4f8fd2083dc88e34a2af2 wallets 0x04453fba30babcb009a8cb821d5593b7a70839f0, 0xee43e3154002a54639e4bf0f362f75f98dc81cc2, 0x0004a2907dff791c5150691db84cd10f9b31369b, 0x3bc46589dbc78c35b7fb0f5fb532c628c3486d04
  • on-chaintoken.bundles: block 105273935 commonFunder 0xcdd37ada79f589c15bd4f8fd2083dc88e34a2af2 pctSupply 1.2898; block 105273963 pctSupply 14.6232

  Confirmed by: facts, claude-opus-5.5, grok-4.7, gpt-5.6-sol

• MediumOne funder's 5 wallets bundled 6.14% of supplysnipersDeterministic

  5 wallets funded by the same address 0xf70d…dbef bought 6.14% of the supply together (block 105273936). Spreading one buyer over many wallets hides how much a single entity controls.

  Evidence

  • on-chainbundles with commonFunder 0xf70da97812cb96acdf810712aa562db8dfa3dbef: 5 wallets, pctSupply 6.14

  Confirmed by: facts

• MediumPosted 1.3% trading fee contradicts the measured 2.65% feeclaims-mismatchAI

  Fork simulation measured 2.65% on both buys and sells, not 1.3%. The on-chain configuration consists of a 1% protocol fee plus a 1.65% creator tax, and the recorded protocol share is 30%, not the claimed 25%.

  Evidence

  • Xhttps://x.com/agentic_pad/status/2117689688626790679
    Show source

    Traders pay 1.3%. Exactly 1% of every trade lands in the agent's wallet and splits 50 / 25 / 25: half stays with the agent, a quarter to the creator, a quarter to the platform.

  • forkbuyTaxPct = 2.65; sellTaxPct = 2.65
    Show source

    curve fees: feeBps=100, creatorTaxBps=165 — buy/sell "tax" above includes these curve fees (relative to fee-less constant-product price)

  • on-chainPonsV2BondingCurve.feeBps = 100 (1%)
  • on-chainPonsV2BondingCurve.creatorTaxBps = 165 (1.65%)
  • on-chainPonsV2BondingCurve.protocolFeeShareBps = 3000 (30%)

  Confirmed by: gpt-5.6-sol · Disputed by: claude-opus-5.5

• MediumOnly about $172 of real liquidity left in the bonding curveliquidityAI

  After the round-trip trading, the curve's real quote reserve is about 0.0709 ETH, reported as $172 of liquidity. The curve's pricing adds 1.68 ETH of phantom (virtual) reserve on top of that real balance. Graduation needs 4.2 ETH. Any sell large relative to the real reserve moves the price sharply, and the token is far from migrating to a Uniswap v4 pool.

  Evidence

  • on-chaincurve realQuoteReserve = 70857674283322621; phantomQuote = 1680000000000000000; graduationThreshold = 4200000000000000000
  • on-chaintoken.liquidity pons-v2-bonding-curve liquidityUsd 172; graduated=false

  Confirmed by: claude-opus-5.5, grok-4.7, gpt-5.6-sol

• MediumPons owner can rescue curve fees and graduation fundsownershipAI

  The launchpad factory owner (0x263e…19Dd) controls several paths that touch curve funds. It can call rescueCurveFees, which triggers the curve's rescueFees, a function that sends ETH. It can call forceSweptGraduation. It can call rescueSweptGraduation, which sends tokens and ETH to a recipient it chooses; a 7-day GRADUATION_RESCUE_DELAY constant exists. It can change the graduationExecutor and the launch deployer helper, and it can override the creator-fee recipient, which has a 3-day timelock. The same address owns the buyback vault, the LP locker and the router. The curve and factory source were not provided, so the exact limits are unverified: whether rescueFees only touches accrued fees, and whether rescueSweptGraduation can only run after a failed graduation. This is protocol-wide trust in Pons, not a power specific to this token.

  Evidence

  • on-chainfactory fn rescueSweptGraduation(address,address) [owner] @PonsV2LaunchFactory.sol:L1254-L1286 — sends tokens out, sends ETH
  • on-chainfactory fn rescueCurveFees(address) [owner]; curve fn rescueFees() [factory] — sends ETH
  • on-chainfactory param GRADUATION_RESCUE_DELAY = 604800; CREATOR_FEE_RECIPIENT_TIMELOCK = 259200; owner = 0x263ed295dAFaE1d9AAdD6E56c4B6F9f38eE019Dd

  Confirmed by: claude-opus-5.5, grok-4.7 · Disputed by: gpt-5.6-sol

• Low'Dev supply is locked' claim while deployer holds 1.24% unlockedclaims-mismatchAI

  Although a locker holds 2% of supply, the deployer itself currently holds another 1.2434%. The categorical statement that dev supply is locked is therefore misleading because the identified dev wallet still controls transferable tokens.

  Evidence

  • Xhttps://x.com/agentic_pad/status/2117687918567600263
    Show source

    Dev supply is locked.

  • on-chainDeployer 0x54fca90e647aed824fc47d0b968dbf132e7cdd24 holds 1.2434% now
  • on-chain0xbe0b139abc90723af76a89d3051f60ba1b64c8d9 (PinkLock02 (token/LP locker, labelled on Blockscout)) 2%

  Confirmed by: gpt-5.6-sol · Disputed by: claude-opus-5.5

• LowOwner can raise the hook fee before this token graduatesfeesAI

  After graduation, trading happens through PonsV2MemeHook. Its owner (0x263e…19Dd) can change the global policy: hook fee up to 10%, protocol share up to 50%, buyback share, fee recipient and sweep operator. Each pool snapshots these terms when it is registered and keeps them for life. Creator tax plus hook fee is capped at 20% in total. The fees taken in afterSwap go into pending buckets split between protocol, creator and buyback vault. rescuePoolFees pays that same split, not the owner. Swaps cannot be blocked by owner settings. The remaining risk is that the owner raises the hook fee (up to 10%) before this token graduates, so the graduated pool would carry the higher fee. The owner also chooses the sweep operator, whose minimum-output values decide the price of fee conversions.

  Evidence

  • codePonsV2MemeHook.sol:L239-L278
    Show source

    function setProtocolFeeShareBps(uint256 bps) external onlyOwner {
            if (bps > MAX_PROTOCOL_FEE_SHARE_BPS) revert InvalidBps();
            protocolFeeShareBps = bps;
            emit ProtocolFeeShareUpdated(bps);
        }
    
        function setBuybackBurnBps(uint256 bps) external onlyOwner {
            if (bps > BASIS_POINTS) revert InvalidBps();
            buybackBurnBps = bps;
            emit BuybackBurnBpsUpdated(bps);
        }
    
        function setHookFeeBps(uint256 bps) external onlyOwner {
            if (bps > MAX_HOOK_FEE_BPS) revert InvalidBps();
            hookFeeBps = bps;

  • codePonsV2MemeHook.sol:L342-L352
    Show source

    if (
                policy.protocolFeeRecipient == address(0) || policy.protocolFeeShareBps > MAX_PROTOCOL_FEE_SHARE_BPS
                    || policy.buybackBurnBps > BASIS_POINTS || policy.hookFeeBps > MAX_HOOK_FEE_BPS
                    || policy.maxInternalPriceImpactBps == 0 || policy.maxInternalPriceImpactBps >= BASIS_POINTS
            ) {
                revert InvalidBps();
            }
            // The curve bounds the same sum in its own constructor rather than
            // inheriting it from the factory. A pool taking more than the whole
            // unspecified leg would flip the swapper's output delta negative.
            if (uint256(creatorTaxBps) + policy.hookFeeBps > MAX_TOTAL_TRADE_FEE_BPS) revert InvalidBps();

  • codePonsV2MemeHook.sol:L444-L485
    Show source

    function _afterSwap(address, PoolKey calldata key, SwapParams calldata params, BalanceDelta delta, bytes calldata)
            internal
            override
            returns (bytes4, int128)
        {
            // The conversion and buyback legs swap against this same pool, but
            // they are never taxed here: v4-core skips a pool's hooks when the
            // hook itself is the caller (Hooks.afterSwap). Were that not so, the
            // skim would come out of the buyback before it reached the vault.
            PoolId poolId = key.toId();
            LaunchInfo memory info = launches[poolId];
            if (!info.registered) return (IHooks.afterSwap.selector, 0);
            if (info.hookFeeBps == 0 && info.creatorTaxBps == 0) return (IHooks.afterSwap.selector, 0);
    
            bool specifiedIsCurrency0 = (params.amountSpecified < 0) == params.zeroForOne;

  • codePonsV2MemeHook.sol:L580-L603
    Show source

    function _rescueCurrency(PoolId poolId, LaunchInfo memory info, address currency)
            private
            returns (uint256 protocolAmount, uint256 creatorAmount)
        {
            uint256 total = pendingFees[poolId][currency];
            uint256 tax = pendingCreatorTax[poolId][currency];
            if (total == 0 && tax == 0) return (0, 0);
    
            pendingFees[poolId][currency] = 0;
            pendingCreatorTax[poolId][currency] = 0;
            // The rescue pays the creator their whole bucket rather than running
            // a buyback, since the swap would fail for the same reason the escrow
            // path did. Clearing the earmark keeps it from surviving as a claim
            // on fees this call has already paid out.
            pendingBuyback[poolId][currency] = 0;

  Confirmed by: claude-opus-5.5

• LowWebsite is 4 days old and its dApp test did not finishwebsiteAI

  agenticpad.family was registered on 2026-10-31, four days before analysis. The fetched site had only 172 characters of text and no Wayback Machine snapshots. The automated dApp test stopped after 5 minutes without finishing, so the site's on-chain interactions are unverified.

  Evidence

  • on-chainsite.domainCreatedAt = 2026-10-31T01:03:52.497Z; waybackSnapshots = []
  • forkwarnings: dApp tests stopped after 5 min (the site did not finish responding) — skipped

  Confirmed by: claude-opus-5.5, grok-4.7, gpt-5.6-sol

Positive signals (1)

• PositiveToken is a fixed-supply template with no owner powersotherAI

  PonsV2LauncherToken is OpenZeppelin ERC20 plus ERC20Burnable. The whole supply is minted once, to the bonding curve, in the constructor. There is no owner, no mint after deploy, no fee, no blacklist, no pause and no upgrade path. deployer is stored only as reference data. Nobody can change balances or block transfers at the token level. burnFrom only works with an allowance the holder gave themselves.

  Evidence

  • codePonsV2LauncherToken.sol:L44-L71
    Show source

    constructor(
            string memory name_,
            string memory symbol_,
            string memory logo_,
            string memory description_,
            Socials memory socials_,
            address deployer_,
            address curve_,
            address launchFactory_,
            uint256 supply_
        ) ERC20(name_, symbol_) {
            if (deployer_ == address(0) || curve_ == address(0) || launchFactory_ == address(0)) {
                revert ZeroAddress();
            }

  • codeERC20.sol:L176-L204
    Show source

    function _update(address from, address to, uint256 value) internal virtual {
            if (from == address(0)) {
                // Overflow check required: The rest of the code assumes that totalSupply never overflows
                _totalSupply += value;
            } else {
                uint256 fromBalance = _balances[from];
                if (fromBalance < value) {
                    revert ERC20InsufficientBalance(from, fromBalance, value);
                }
                unchecked {
                    // Overflow not possible: value <= fromBalance <= totalSupply.
                    _balances[from] = fromBalance - value;
                }
            }

  • on-chaincontract.privileges = [] ; isProxy = false
  • staticslither 0.11.6: issues = []

  Confirmed by: claude-opus-5.5, grok-4.7 · Disputed by: gpt-5.6-sol

Notes (5)

• InfoBuys and sells work with a fixed 2.65% curve feefeesAI

  The fork simulation bought and sold on the bonding curve. A second wallet that received tokens by transfer could also sell, and a sell one hour later worked. Plain transfers are not taxed. The 2.65% on buys and sells is the curve's 1% protocol fee plus a 1.65% creator tax. Both are stored as immutable values in the curve, so nobody can raise them for this token. The launch snipe tax, which started at 99%, decayed to 0 after 3 seconds.

  Evidence

  • forksimulation: canBuy=true, canSell=true, buyTaxPct=2.65, sellTaxPct=2.65, transferTaxPct=0
    Show source

    second wallet (received via transfer) could sell

  • on-chaincurve feeBps = 100 (immutable), creatorTaxBps = 165 (immutable)
  • on-chainlaunchpad notes: snipeTaxNow=0bps over 3s

  Confirmed by: claude-opus-5.5, grok-4.7, gpt-5.6-sol

• InfoUniswap v4 aggregator hook is ownerless and cannot block sellsotherAI

  The Uniswap v4 pool 0x88b18bd3… uses PonsCurveAggregator. It is a singleton with no owner. It passes swaps straight to the bonding curve, using the curve's own fees, and the pool fee is fixed at 0. It refuses exact-output swaps, buys that would only partly fill, and any trade once the curve is ready to graduate. Selling directly on the curve still works in those cases. It has no admin settings that could block sellers.

  Evidence

  • codePonsCurveAggregator.sol:L113-L136
    Show source

    /// @dev Mirrors PonsV2BondingCurve.buy: the fee, creator tax and snipe tax come off the quote side, the rest buys
        ///      tokens at constant product. The snipe tax is the hook's own (the curve charges it per recipient).
        function _quoteBuy(IPonsV2BondingCurve curve, uint256 quoteIn) private view returns (uint256 tokensOut) {
            if (curve.graduated() || curve.readyToGraduate()) revert CurveClosed();
            uint256 feeBps = curve.feeBps();
            uint256 taxBps = curve.creatorTaxBps();
            uint256 snipeBps = curve.currentSnipeTaxBps(address(this));
            if (snipeBps != 0) {
                uint256 maxSnipeBps = BPS - feeBps - taxBps - 100;
                if (snipeBps > maxSnipeBps) snipeBps = maxSnipeBps;
            }
            uint256 net = quoteIn - (quoteIn * feeBps) / BPS - (quoteIn * taxBps) / BPS - (quoteIn * snipeBps) / BPS;
            (uint256 quoteReserve, uint256 tokenReserve) = curve.getReserves();
            tokensOut = PonsCurveMath.getAmountOut(net, quoteReserve, tokenReserve);
            if (tokensOut > curve.sellableTokens()) revert WouldPartiallyFill();

  • codePonsCurveAggregator.sol:L152-L179
    Show source

    function _conductSwap(Currency settleCurrency, Currency takeCurrency, SwapParams calldata params, PoolId poolId)
            internal
            override
            returns (uint256 amountSettle, uint256 amountTake, bool hasSettled)
        {
            if (params.amountSpecified >= 0) revert ExactOutputNotSupported();
            Market memory m = _market(poolId);
            amountTake = uint256(-params.amountSpecified);
            poolManager.take(takeCurrency, address(this), amountTake);
    
            if (_isBuy(m, params.zeroForOne)) {
                // refuse a partial fill up front: the curve would refund the rest, which a v4 exact-input swap cannot
                uint256 expected = _quoteBuy(m.curve, amountTake);
                uint256 quoteBefore = takeCurrency.balanceOfSelf() - amountTake;
                if (takeCurrency.isAddressZero()) {

  Confirmed by: claude-opus-5.5

• Info2% of supply in a PinkLock02 locker with unknown unlock dateholdersAI

  The PinkLock02 locker holds 2% of supply. Its beneficiary and unlock date are not in the material, so it is unknown whose tokens these are and when they unlock. Excluding the curve and the locker, holder concentration is low: the top 10 hold 2.047% and the deployer holds 1.2434%.

  Evidence

  • on-chaintopHolders: 0xBE0b139ABc90723Af76a89D3051f60Ba1B64c8d9 pct 2 label PinkLock02
  • on-chaintoken.top10Pct = 2.047; deployer 1.2434%

  Confirmed by: claude-opus-5.5, grok-4.7, gpt-5.6-sol

• InfoScanner bot numbers differ from our holder dataclaims-mismatchAI

  Automated scanner and launch-alert accounts called $AP sketchy: snipers still holding, a launch crew that sold, dozens of holders, and (in one bot) the same X account tied to other launches. Those numbers are not confirmed here. The basics we were given show 14 holders, the top 10 holding about 2% combined, and no prior launches for this deployer. Treat the bot scores as allegations, not as proof the coin was dumped or that the team has a rug history.

  Evidence

  • Xhttps://x.com/Larpcheckbot/status/2117700522232762431
    Show source

    AP · TOTAL 43/100 · NARRATIVE: PARTLY REAL 45 · COIN: SKETCHY 40 Real-looking platform, sketchy coin: launch crew dumped 54%, top 10 hold 22%. Red flags: - Snipers hold 13%; launch crew sold 54%

  • Xhttps://x.com/Tetrascoreio/status/2117696364683665584
    Show source

    AP · Safety D 46 🟠 🔴 13 early wallets bought 60.6% of supply in the first minute and sold… 🟠 Same X account as 3 other launches (copy, relaunch or migration) 👥 60 holders · dev: 1 launch, 0 grad

  • Xhttps://x.com/LongLaunches/status/2117703668036161608
    Show source

    MC $19.0K · Vol 24h $25.6K · Buyers 161 · Holders 78 Bonding curve 77% filled ⚠️ Snipers 20% · Bundlers 13% · Insiders 2.0% · Creator 0.5%

  • on-chainholdersCount: 14; top10Pct: 2.047; deployerHistory: []; deployer = 0x54FCA90E647aed824fc47D0b968DBf132E7CdD24
    Show source

    holdersCount 14, top10Pct 2.047, deployerHistory empty

  Confirmed by: grok-4.7, gpt-5.6-sol · Disputed by: claude-opus-5.5

• InfoX account and site are only days old, with an anonymous teamsocialAI

  @agentic_pad was created on 1 Nov 2026 and the site domain on 31 Oct 2026. The token was deployed on 3 Nov 2026. The account has 29 followers and its visible posts are all from launch day, mostly product claims (1% of trades to an agent wallet, hourly payouts, dev supply locked, creator seat selling soon). No named person is identified. That fits a brand-new launchpad token, not a hidden old project, but the payout, lock, and 'nobody can touch the treasury' lines are posts only — the site is one sentence and does not prove them.

  Evidence

  • Xhttps://x.com/agentic_pad/status/2117687918567600263
    Show source

    Dev supply is locked.

  • Xhttps://x.com/agentic_pad/status/2117689688626790679
    Show source

    Nobody can buy in between, and the dev team never holds the agent's money.

  • webhttps://agenticpad.family/
    Show source

    Agentic Pad: every token launched here gets its own AI agent on Robinhood Chain. 1% of every trade funds the agent; it buys back, airdrops to holders and trades on its own.

  • on-chaindeployedAt: 2026-11-03T18:34:54.000Z
    Show source

    X createdAt 2026-11-01T14:48:13.000Z; domainCreatedAt 2026-10-31T01:03:52.497Z; followers 29

  Confirmed by: grok-4.7, gpt-5.6-sol