[ Coin launchpad · Robinhood Chain · pre-launch ]

Every sale is a burn.

Launch a coin that buys an NFT floor, shelves it at a markup, and burns itself every time the NFT leaves.

How the loop works Read the docs
Keeper
None
Withdraw fn
None
Exits
Both burn
Runs onRobinhood ChainPons V2Seaport 1.6Uniswap v4

Shelf · $COIN ⇄ one Robinhood Chain collection

Sample shelf · pre-launch

Bid · simulated clock

next +2% in 2:00

0.0235 ETH

cap 0.0630 · reaches it in 7h 42m if nobody sells

  • #871 · day 2

    cost 0.0420 → shelf 0.0504 ETH · pull 0.0462 ETH

    pull = burn 1.54M $COIN (0.15% of supply)

    0.050 ETHpull 1.1×

  • #154 · day 9 · decaying

    cost 0.0400 → shelf 0.0472 ETH · pull 0.0432 ETH

    pull = burn 1.44M $COIN (0.14% of supply)

    0.047 ETHpull 1.1×

  • 8 empty slots · the shelf keeps bidding until it holds 10

ETH on shelf 0.118 · withdraw none · owner none

every figure computed from the shelf formulas

This is the product's own shelf view, on sample figures. The real shelf opens with the first coin launched through Firesale; nothing is live and nothing has been burned yet.

Firesale · $FIRESALE · contract address

Pre-launch

To be announced.

The token launches through Firesale itself, after the contracts are verified on-chain, so there is nothing to buy yet. This receipt is where the address will be printed. Until it is, any contract claiming to be Firesale is a fake.

Verify the address against this page and the project's own channels before sending anything.

[ The loop ]

Tag it. Sell it.
Burn it.

A coin, one collection, and a shelf with no owner. Fees go in; NFTs come and go; the coin only ever gets smaller.
  1. 01

    Trade

    Anyone trades the coin on the curve, later on the pool. Each trade pays a 1% fee plus the creator tax the launcher set (0–10%).

  2. 02

    Route

    Every wei reaching the coin's router splits 80% to its shelf, 10% to a buyback of $FIRESALE, 10% to ops. The split is a constant in bytecode.

  3. 03

    Bid

    The shelf's bid for its collection starts low and climbs 2% of the last fill every ten minutes until a holder sells or a listing fills.

  4. 04

    Shelve

    The NFT is tagged at 1.2× cost. Holders of the coin can pull it at 1.1× cost in coins at any time — always 0.1× cheaper than an ETH buyer.

  5. 05

    Burn

    ETH from a sale buys the coin back and burns it. A pull burns the coins on the spot. Either way, supply falls and every remaining holder's share rises.

Where the coin ends up

exit A → ETH → buyback → burn  ·  exit B → coins → burn

[ The rising bid · interactive ]

A price without
a keeper.

After each fill the bid restarts at half the last price and climbs 2% of it every ten minutes. The cheapest seller in a collection is always the first to accept a bid that only goes up. Move the sliders — these are the shelf's real formulas.

[ Inputs ]

bid(t) = min(cap, 0.5·L + 0.02·L·⌊t/600s⌋) · cap = min(shelfETH, 1.5·L)
shelf = cost · max(0.9, 1.2 − 0.01·max(0, day − 7)) · pull = shelf − 0.1·cost

The bid right now

cap in 7h 42m

0.0235 ETH

starts at 0.0210 · cap 0.0630 · capped at 1.5× the last fill

Exit A · buy with ETH

0.0504 ETH

1.20× cost

All of it buys the coin back and burns it, in chunks of ≤0.25 ETH.

Exit B · pull with coins

1.54M coins

worth 0.0462 ETH · 0.15% of supply

Burned on the spot. Always 0.1× cost cheaper than the ETH price.

Sell to shelf

any holder of the collection sells at the current bid in one transaction

Fill a listing

anyone executes a Seaport listing priced at or below the bid and keeps a 0.5% tip

Stale rule

from day 7 both prices fall 1% of cost a day, to 0.9× / 0.8× — a discounted exit still burns; a frozen shelf burns nothing

[ Two exits ]

Both burn.
Every time.

An NFT leaves a shelf in exactly two ways. Each produces a receipt — an image with the burn, the share of supply and the transaction on it. A receipt cannot be made for an event that never happened.

Exit A

Buy it with ETH

1.2× cost

  • ◆100% of the ETH enters the burn bucket
  • ◆burnTick() buys the coin back in chunks of ≤0.25 ETH every 10 minutes and burns what it buys
  • ◆per 0.80 ETH of shelf spend → 0.96 ETH of coin bought back and burned

Burn receipt

Sample

SOLD

collection item #3038

burned
0.61 ETH → buyback
of supply
in progress, tick by tick
cost → exit
0.042 → 0.050 ETH
how
ETH → buyback + burn
tx
0x… (sample — no transaction yet)

Exit B

Pull it with coins

1.1× cost

  • ◆burn coins worth 1.1× cost, take the NFT
  • ◆always 0.1× cheaper than the ETH price — the holder's edge
  • ◆coins are priced at the lower of spot and the 30-minute average, so a pump never makes a pull cheaper

Burn receipt

Sample

PULLED

collection item #871

burned
1,900,000 coins
of supply
0.19%
cost → exit
0.042 → 0.046 ETH in coins
how
coins burned on the spot
holder tier
Ember
tx
0x… (sample — no transaction yet)

The coin only gets smaller

Simulation · sample numbers · nothing is live

supply

1B of 1B

burned so far
0 coins
of supply
0.00%

Coin priced at a 30 ETH market cap. Every figure comes from the shelf formulas; a sale's buyback is shown as one step here, on-chain it drips in ≤0.25 ETH ticks. The real receipts start with the first shelf.

Exit feed

  • waiting for the first exit…

[ The shelf ]

Rules in bytecode,
not in a wallet.

A shelf is a contract with no owner and no way out for its ETH except an NFT. What it holds, what it can do and what it cannot are fixed when it is cloned.

Holds

  • ◆ETH: the bid budget, plus a separate burn bucket
  • ◆up to 10 NFTs from exactly one collection

Can

  • ◆claim its fees from the launchpad escrow
  • ◆bid for, and buy, one floor NFT at a time
  • ◆sell a shelf NFT at the shelf price
  • ◆release a shelf NFT to anyone who burns the pull price
  • ◆buy its own coin back and burn it

Cannot

  • ✕withdraw ETH to any address
  • ✕move an NFT except through a sale or a pull
  • ✕change its collection, its split or its prices
  • ✕be paused, upgraded or re-owned by anyone, including us

Split, forever

80 / 10 / 10

shelf / $FIRESALE burn / ops · a constant in the bytecode

What Firesale will not do

  • ✕no random draws, no tickets, no single-holder handouts
  • ✕no keeper wallet that can move shelf ETH
  • ✕no off-chain purchases, no bridged collections (v1)
  • ✕no withdraw function in any contract
  • ✕no admin key on shelves, no pause switch
  • ✕no discount or priority for holding $FIRESALE
  • ✕no cut of NFT sales, pulls or buybacks

[ Why ]

The first launchpad
leaks value.

Fee-funded NFT sweeps proved demand on Robinhood Chain: in its first 4 days the category leader launched 229 coins, routed about $355K of creator fees and bought 2,328 NFTs. Its design leaks in four places.
  1. 01

    No loop

    The NFT is given away or destroyed. The coin whose trading paid for it receives nothing back.

  2. 02

    Trust gaps

    About 87% of vault ETH sits in keeper-run vaults for collections on other chains, bought off-chain. Core contracts were unverified at review.

  3. 03

    Blue-chip theatre

    Famous collections sell the story, yet ~97% of NFTs bought were floors under 0.03 ETH, and not a single CryptoPunk was bought.

  4. 04

    Ticket payoff

    One NFT spread across thousands of holders is a ticket, not a mechanism.

Them vs us · figures as of 2026-09-23

ThemNFT-sweep launchpads hand each NFT to one holder picked at random

UsEvery NFT that leaves the shelf burns the coin that paid for it

ThemThe coin gets nothing back

UsMarkup + burn close the loop

ThemMost vault ETH buys NFTs on other chains, off-chain, through a keeper wallet

UsRobinhood Chain only: every buy settles on-chain in the same contract

ThemA keeper posts a price cap

UsA rising bid sets the price; no keeper, no oracle

ThemCore contracts unverified

UsVerified source from block 1

ThemHolders hold a ticket

UsHolders can pull any NFT at 1.1× cost, paid in coins

ThemPlatform token's own fees go to a team wallet

Us10% of every coin's fees buy and burn $FIRESALE

[ Where the money goes ]

One split.
No exceptions.

Every wei that reaches a coin's router is divided the same way. Firesale's only income is its 10% share; buyers' ETH and pulled coins are burned, never collected.

Router split

  • Shelf · buys the collection's floor

    80%

  • $FIRESALE buyback + burn · from every coin, forever

    10%

  • Ops · reviews, bounty, indexer, bot fees

    10%

Per 1 ETH traded · creator tax 1%

router
0.0170 ETH
launchpad
0.0030 ETH
→ shelf
0.0136 ETH
→ $FIRESALE burn
0.0017 ETH
→ ops
0.0017 ETH

Per 1 ETH reaching a router

  • 0.80 → shelf buys an NFT at 0.80
  • sold for ETH at 1.2× → 0.96 ETH buyback + burn
  • pulled at 1.1× → 0.88 ETH of coin burned
  • 0.10 → buy + burn $FIRESALE
  • 0.10 → ops: reviews, bounty, indexer, bot fees

What Firesale never takes

  • ✕no launch fee on top of the launchpad's
  • ✕no cut of NFT sales from shelves
  • ✕no fee on pulls
  • ✕no share of buybacks
  • ✕no access to shelf ETH, ever

Scenarios · benchmark arithmetic, not forecasts

benchmark: ~32 ETH/day of router fees, 2026-09-23 · ETH $2,690

1% of that flow

0.32 ETH/day

to shelves
0.26 ETH
$FIRESALE burn
0.032 ETH
ops
$86

10% of that flow

3.20 ETH/day

to shelves
2.56 ETH
$FIRESALE burn
0.32 ETH
ops
$860

25% of that flow

8.00 ETH/day

to shelves
6.4 ETH
$FIRESALE burn
0.8 ETH
ops
$2,150

Real flow is power-law: the top 10 coins made 74% of the benchmark's fees. Running costs of a few hundred dollars a month are covered at roughly $120K of monthly volume.

[ $FIRESALE ]

The platform token
eats its own cooking.

$FIRESALE launches through Firesale like any coin and runs the same loop on its own shelf. On top of that, 10% of every router on the platform buys and burns it.

With vs without

Without

  • ◆ Launch coins
  • ◆ Trade, pull, buy, sell to shelf
  • ◆ Every receipt and alert
  • ◆ Full docs and data
  • ◆ Everything a holder needs

With $FIRESALE

  • ◆ Same product, same prices
  • ◆ Tier badge on every receipt
  • ◆ Weekly votes on the next collections to allowlist
  • ◆ Ember+ nominate collections
  • ◆ Exposure to the platform burn from every coin

$FIRESALE never buys a discount, a better price, priority on a shelf, or first access to an NFT. The product is identical for everyone. The token is identity, curation and the platform-wide burn.

Platform burn flywheel

  1. 1more coins launched on Firesalethen
  2. 2more trading fees through every routerthen
  3. 310% of each router buys and burns $FIRESALEthen
  4. 4supply falls with total platform volumethen
  5. 5attention on the platform token → more launches

→ loops back to 1. Two burn engines, one token: its own shelf plus a slice of everyone else's.

Visitor

hold 0

  • ✓ launch, trade, pull, buy, sell to shelf
  • ✓ every receipt, alert and page

Spark

hold 500K · 0.05% of supply

  • ✓ Spark badge on receipts
  • ✓ 1 collection vote per week

Ember

hold 2.5M · 0.25% of supply

  • ✓ Ember badge on receipts
  • ✓ 3 collection votes per week
  • ✓ nominate a collection for review

Blaze

hold 10M · 1% of supply

  • ✓ Blaze badge on receipts
  • ✓ 10 collection votes per week
  • ✓ one featured-shelf request a month (curated, never sold)

Tiers buy badges and curation votes. They never buy a discount, priority or first access — the product is identical at every tier.

Supply1,000,000,000 $FIRESALE
Venuelaunched through Firesale, on Pons V2
Curvebonding curve → graduates at 4.2 ETH into a Uniswap v4 pool
Paired withone Robinhood Chain collection, announced at launch
Creator tax1%, routed like every coin: 80 / 10 / 10
Presalenone
Team allocationnone — any team buy happens on the public curve and is disclosed with the address
Vestingnone (nothing to vest)
Burnsreal ERC-20 burns: totalSupply falls and every receipt links the transaction

[ Contracts ]

Four contracts,
no admin on the shelf.

Verified source from the first block. Addresses are published here when they deploy; until then any address claiming to be Firesale is not.

  • FiresaleShelf

    one per coin · no owner, no withdraw, no upgrade

  • FiresaleLauncher

    clones a shelf, launches the coin with the shelf as fee recipient

  • CollectionRegistry

    allowlist · multisig + 48h timelock · future launches only

  • PlatformBurner

    collects the 10% slices, buys and burns $FIRESALE

  • Pons V2

    bonding curve + fee escrow (external)

  • Seaport 1.6

    buy and sell NFTs (external)

  • Uniswap v4

    post-graduation pool, buybacks (external)

  • Indexer + app

    reads events, renders receipts, no special rights

harvest()claim escrow, split 80/10/10
currentBid()the rising-bid formula, view
sellToShelf(tokenId)a holder sells at the bid
fillListing(order)execute a Seaport listing ≤ bid, 0.5% tip
buy(tokenId)exit A at the shelf price
pull(tokenId, maxCoins)exit B, burns coins
burnTick()next buyback chunk + burn
poke()apply decay, refresh the 30-minute average

Phase 0 · proofs on a mainnet fork before the final contracts

  • provena contract can be both deployer and fee recipient on Pons V2
  • to provea shelf fills a Seaport 1.6 listing on Robinhood Chain
  • to provecurve buy + burn from a contract, before graduation
  • to proveUniswap v4 buyback + burn, after graduation
  • to provepost-graduation fee sweep without the operator
  • to provehow launch-block anti-snipe tax is routed

Each proof is a fork test that stays in the repository.

[ Security ]

Attack, defence,
and what we cannot fix.

AttackOverpriced or fake floor sold into the shelf

Defencethe bid starts low and only rises; the cheapest seller wins; a cap per fill; allowlisted collections only

AttackSandwich around the buyback

Defencesmall chunks, spaced ticks, slippage bound from the 30-minute average

AttackPump the coin to make a pull cheaper

Defencepull priced at the lower of spot and the 30-minute average

AttackMalicious NFT contract (hooks, reentrancy)

Defencestandard ERC-721 only, allowlisted; reentrancy guards; checks-effects-interactions on every transfer

AttackGriefing the shelf with dust NFTs

Defenceonly the paired collection; a 10-slot cap

AttackLaunchpad owner reroutes a coin's future fees

Defencecannot be prevented (3-day timelock on the launchpad); monitored and announced publicly; ETH and NFTs already on a shelf stay where they are

AttackDeployer key compromise

Defencefresh keys, never reused test keys; nothing to steal after deploy — no admin powers on shelves

Trust points we cannot remove

  • !Pons V2's owner can set a new creator-fee recipient for any coin after a 3-day timelock. Future fees can move; what already reached a shelf cannot.
  • !After graduation, some pool-fee conversions may need the launchpad's operator to settle.
  • !The launchpad can switch new launches off.
  • !Our Registry multisig decides which collections are offered for future launches. It cannot touch a shelf.

We would rather write these down than have you find them.

[ Roadmap ]

Proof first.
Token last.

  1. Phase 0 · week 1

    now

    Proof

    • ◆ fork tests: Seaport fill, curve + v4 buyback/burn, post-graduation sweep, anti-snipe routing
    • ◆ fresh deployer keys, multisig for the Registry
    • ◆ security review scope written before code freeze

  2. Phase 1 · weeks 2–4

    MVP

    • ◆ Launcher, Shelf, Registry, PlatformBurner verified on mainnet, with conservative v1 caps
    • ◆ launch flow, shelves, coin pages, buy / pull, receipts, docs
    • ◆ 3–5 allowlisted Robinhood Chain collections
    • ◆ maintenance bot live (no special rights)
    • ◆ $FIRESALE launched through Firesale itself

  3. Phase 2 · months 2–3

    Growth

    • ◆ tiers, badges and weekly collection votes
    • ◆ Shelf Watch leaderboards + shelf alerts
    • ◆ collection pages with sell-to-shelf for NFT holders
    • ◆ external security review, then cap raises
    • ◆ bug bounty

  4. Phase 3 · months 4–6

    Scale

    • ◆ shelf NFTs listed on public marketplaces through contract signatures
    • ◆ creator dashboard: fees, fills, burns, holders
    • ◆ embeddable floor-bid widget for NFT project sites
    • ◆ more collections as the chain's NFT market grows

  5. Phase 4 · 6–12 months

    Ecosystem

    • ◆ verified collection coins: a collection's own team endorses one coin
    • ◆ multi-collection shelves for curated sets
    • ◆ other collections only if every buy settles on-chain in the shelf itself — no keeper custody, ever

[ Decisions ]

What we chose,
and why.

v1 parameters. Confirmed before the contracts are final; anything that changes will be changed here first.

Router split 80 shelf / 10 $FIRESALE burn / 10 ops, creator 0%

Creators launched 229 coins in 4 days at 0% elsewhere; the 10% burn ties the platform token to all volume

Exit A 1.2×, exit B 1.1×, decay from day 7 to 0.9×

Markup funds extra burn; the pull discount is the holder's edge; decay keeps shelves moving

Robinhood Chain collections only in v1

Every buy on-chain, no keeper custody; costs the blue-chip story, buys trust

$FIRESALE launched through Firesale

Dogfooding beats a team wallet; a cheap floor means frequent cycles and receipts

Rising bid, not a set cap

Price discovery with no keeper and no oracle

One shelf per coin

No attribution disputes between coins sharing a collection

No random draws anywhere

A mechanism, not a prize

Pull priced at min(spot, 30-min average)

Pumping never makes a pull cheaper

Buyback in ≤0.25 ETH ticks

Sandwiching a tick is not worth the trade

Verified contracts + v1 caps

Trust is the differentiator; caps limit a first-version bug

Product live before the token

The site and first shelves exist before $FIRESALE trades

Watched

  • ◆ coins launched / graduated
  • ◆ router fees per day
  • ◆ fills per day; average fill vs cheapest listing
  • ◆ exits per day, split sale vs pull
  • ◆ value burned per day, per coin and platform-wide
  • ◆ stale NFTs (day 7+) as % of shelf inventory
  • ◆ receipts shared

Deliberately ignored

  • ✕ follower counts
  • ✕ market cap on its own
  • ✕ total value locked as a headline
  • ✕ number of collections listed

[ The honest take ]

Strengths, risks,
no marketing.

Strengths

  • ◆Closes the loop the category left open: every exit is a burn, on-chain, per coin
  • ◆No keeper anywhere in the money path; price discovery by a rising bid
  • ◆Verified, ownerless shelves from block 1 — the trust story is the product
  • ◆Cheap network fees make the drip buyback and frequent bid pokes viable
  • ◆Receipts turn every exit into shareable proof with the transaction on it

Risks

  • !Robinhood Chain NFT floors are thin; a shelf can fill with NFTs nobody wants and burn slowly
  • !No blue-chip collections in v1, so less of the hype the incumbent enjoys
  • !Dependence on the launchpad: fee recipients can be rerouted after a timelock, launches can be switched off
  • !Contract risk: shelves hold ETH and NFTs; a review and v1 caps limit, but do not remove, it
  • !A copy is possible; the moat is speed, receipts and never having had a keeper

First mover

The first ownerless buy → relist → burn loop on Robinhood Chain, packaged as a launchpad.

Network effects

Coins fund bids, bids attract NFT communities, those communities trade the coins.

Data

Every fill is a real, on-chain floor print. Over time the cleanest floor history on the chain.

Switching costs

A coin's pairing is permanent and its shelf holds NFTs. Holders cannot move a shelf to a copycat.

Brand

One line, one mark, one receipt format people recognise in the feed.

Technical

Rising bid, pull guard, drip buyback, admin-free shelves — a system a fork must get right on day one.

Every sale is a burn.

Other launchpads give the NFT away. Firesale sells it back and burns your coin.

Read the docs GitHub ↗Follow ↗