What it would take to leave the lab

PegLab, as this repo, cannot deploy on Kaspa mainnet as money. The pool is 2 tKAS. The oracle is one key. If you treat tPEG as a dollar it depegs. That is the proof of concept. The name “PegLab” stays the teaching bench.

What *can* go to mainnet is a successor that copies working objects from other chains — not this toy’s ticker. This file is that map: existing proofs of concept, how they were actually funded, and the full list of what Kaspa still lacks. It is not legal advice and not a raise.

Authority: DOCTRINE.md. Related: STABLES-GUIDE.md (pegs and capital ladder), VISION.md (finished conversation on a phone), KACHAT.md (chat + pay), BATTLE.md (PegLab vs Parker’s GitHub PoC).

TESTNET-10 LAB. NOT USD. DO NOT RAISE MONEY AGAINST tPEG.

1. Call this lab what it is

Shipping PegLab’s current covenant on kaspa: with a 2 KAS pool is how Chainge-class wraps taught Kaspa a depeg. Mainnet does not add reserves.

2. Copy existing proofs of concept (not their brands)

Steal the mechanism. Do not steal the claim “we are USDC.”

Fiat 1:1 (the market’s winner, ~95% of stable value)

PoC: USDT, USDC, PYUSD.

What to copy

• Redemption at par for cash / T-bills is the product.
• Segregated reserves, monthly examined composition (GENIUS), freeze disclosed in the UI.
• The issuer is a company that can be sued.

How they were funded

• Tether: operating company; float on reserves; not a public-goods grant. Do not copy the opacity.
• Circle: ~$1.1B equity over a decade (Coinbase, BlackRock, Fidelity, …). NYSE IPO June 2025 (~$31, later a mania print). GENIUS Act 2025 is the legal floor they already lived by. 2026: OCC national trust path. That is $10M–$50M+ and lawyers first, then a token.
• PayPal PYUSD: a bank/fintech balance sheet, not a Discord raise.

Kaspa takeaway: if the pitch says dollar, you are Circle’s competitor. You need a bank, not a 2 KAS pool. PegLab will never be this.

Crypto-overcollateral + liquidations

PoC: Maker SAI (Dec 2017) → MCD DAI (2019) → Sky USDS; Liquity LUSD (2021) and BOLD (v2).

What to copy

• Lock more collateral than you mint.
• Redemption vs collateral (Liquity: 1 LUSD/BOLD → $1 of ETH) pulls the price up from below. No sister token.
• Fail closed on stale oracles. Caps published. Surplus buffer.
• Liquity v1: immutable core, no admin pause-as-peg. v2: user-set rates; governance only steers *liquidity incentives*, not the dollar.

How they were funded

• Maker: no retail ICO of DAI. MKR sold over time (forum, private, then market). Polychain / a16z ~$12M (Dec 2017) then a16z $15M for ~6% of MKR (Sep 2018). Maker Foundation in Copenhagen wrote code; later handed governance to the DAO. Trail of Bits / early audits *before* scale. Years of SAI on mainnet before multi-collateral.
• Liquity: protocol company (Liquity AG) + LQTY as a fee-claim / incentive token, not a sister-token that mints to defend $1. Security culture: multiple audits, immutability as the product. No “20% yield pays the peg.”

Kaspa takeaway: this is successor C. Budget $2M–$15M: oracles that are not one key, liquidation path, audits, incident process, MM. Do not use PegLab’s admin oracle.

Peg Stability Module (import someone else’s dollar)

PoC: Maker/Sky PSM — USDS ↔ USDC at $1 until the ceiling.

What to copy

• Arb against a redeemable upstream dollar.
• Publish the cap. When the cap binds, the peg gaps (DAI ~$0.89 in March 2023 because the PSM *was* USDC, and SVB froze Circle).

How it was funded

• Not a separate raise. It piggybacked Maker’s existing MKR/DAI machine and Circle’s reserves.

Kaspa takeaway: successor B then a PSM. Do not raise to *be* the PSM until a named USDC/USDT (or licensed native) is on Kaspa. Chainge-class wraps are issuer plus bridge risk — show that in the UI.

Hard redemption, no PSM

PoC: Liquity. Anyone can redeem the stable for $1 of ETH.

What to copy

• The arb is on-chain. No committee “defends” the peg with a new token.
• Can overpeg in a flight to safety. That is honest.

Kaspa takeaway: a KAS-receipt is even simpler: redeem for sompi, not for $1 of KAS. No oracle. That is successor A. Closest to what Toccata already is.

Receipts / programmable cash (not a dollar)

PoC: Bitcoin DLCs and Taproot escrow; Ethereum “vesting / streaming” (Sablier, Superfluid) as *time-locked units*; local-settlement IOUs that never claim USD; Kaspa-explained backed-receipt (sponsor-fee, no skim of principal).

What to copy

• The unit is the locked asset. Display FX is a banner, not a peg.
• Timeout reclaim (they claim now, or you reclaim after tx.time) — KaChat already sketched this as KaChatPayTimeout.sil.
• Fees paid by a sponsor, not stolen from the receipt.

How they were funded

• Protocol grants, company R&D, or “sell software.” Not a stablecoin sale.

Kaspa takeaway: this is the first mainnet that is allowed. PegLab’s teaching UI becomes a receipt explorer, not a dollar.

What not to copy (funded, then died)

3. How the ones that lived actually got money

Copy the instrument, not the tweet.

Rules for PegLab / successor fundraising

1. Name matches backing. “Receipt” → grants / self-fund. “Dollar” → equity in a licensed company.
2. No tPEG sale. No points. No sister token that mints to defend $1.
3. Grants pay infrastructure (indexer, SilverScript, KaChat signer, postage). They do not buy a peg.
4. If a lawyer says you need a license, stop the token until that is true.
5. If the deck needs this lab’s UI without WILL DEPEG, the deck is the bug.

Order-of-magnitude (same ladder as STABLES-GUIDE):

• Lab: $0–$300k, already built, will depeg
• Receipt on mainnet: $100k–$1M (audit-lite, indexer, wallet copy, KaChat testnet pay)
• Overcollateral protocol: $2M–$15M
• Licensed USD: $10M–$50M+ company

4. Think bigger than the covenant

The chain is the court file. Mainnet fails if any of these are missing.

Product (the scene)

Two people on KaChat phones quote a still number, pay, claim or reclaim. Postage is KAS in the background. No PC KaChat. No $ on a receipt. See VISION.md.

Without that scene, a mainnet series is a block explorer toy.

Protocol (what Toccata is for)

• Live series on Testnet-10 first: genesis outpoint, template hash, series bytes — wallets verify those, not the word tPEG.
• Transfer + redeem accepted by the node, not only src/engine.mjs.
• Kill the admin USD oracle on any asset KaChat would show. Receipts do not need a dollar feed.
• Timeout reclaim (KaChatPayTimeout shape). Split / tip / stake as the same family.
• Sponsor-fee or prepaid grams (KIP-21 mass) so a receiver with 0 KAS still works. Kaspa Core already flagged this.

Wallets and indexers

• KaChat iOS/Android signs the covenant (KasWare-on-the-web is not enough).
• KasWare / Kaspa NG / Kaspium: refuse unknown series.
• Indexer row per address (Kasia or documented equivalent). If the wallet cannot see UNIT UTXOs, the balance is a lie.
• Explorer page: genesis, cap, pause, redeem path.

People, law, ops

• Entity that matches the claim (none for a receipt; company for a dollar).
• GENIUS / MiCA if you say payment stable. Freeze UX named.
• Incident: pause is a brake. Redeem stays open if backing remains. Public post-mortem habit (Liquity/Maker culture, not “trust us”).
• Keys: never in git, never on the public site. Multi-person control if there is an admin at all.
• Support: five failure sentences in the thread (stale, postage, unknown series, paused, timeout).

Distribution

• Phone first (KaChat). Desktop is the lab and docs.
• Merchant QR that opens KaChat with amount filled.
• Agents: quote → 402 → pay. Human optional.
• Other messengers may speak kchat:1:payunit. Do not wait for them.

What Kaspa uniquely adds (so “bigger” is not “become Circle”)

• One-second sequencing, cheap settlement, public evidence.
• Covenants: lock, timeout, split, prove the spend rule — programmable cash.
• Native chat+pay already in a store app (KaChat). Ethereum never had that as a default consumer surface.

What Kaspa does not add: a Treasury, a banking charter, or a reason to ignore the law if you print “USD.”

5. Mainnet checklist (successor A first)

Do not tick “mainnet PegLab.” Tick a receipt (or later a named USD).

Must already be true on Testnet-10

• ☐ One live series; transfer and redeem on-chain
• ☐ Indexer shows UNIT by address
• ☐ KaChat (or a native signer) can pay 1 unit in a 1:1 thread with timeout
• ☐ Postage: pay does not fail because the receiver has 0 KAS
• ☐ Wallets verify genesis / template / series; unknown ticker rejected
• ☐ Copy: no $; WILL DEPEG remains on *this* lab forever
• ☐ AUDIT-LITE + KNOWN-BREAKS closed or accepted; at least one external review for the successor contract

Then mainnet (receipt only)

• ☐ New genesis on kaspa: with published hashes
• ☐ Caps published (backing, supply, dust) — not “infinite”
• ☐ No admin USD oracle on this asset
• ☐ Incident channel and pause policy written *before* first user
• ☐ KaChat review: asset picker KAS | receipt, Child Mode receipts only
• ☐ Still not called USD. Still not a token sale.

Only if you later host a dollar

• ☐ Named issuer in the first sentence
• ☐ Separate asset id from the receipt
• ☐ Freeze / redemption / exam story is *theirs*
• ☐ Company raise if *you* are the issuer — not a PegLab meme round

Never

• ☐ Mainnet this 2 tKAS pool as money
• ☐ List lab tPEG in KaChat 4.x as stablechat
• ☐ Sister token / algo / 20% yield to hold $1
• ☐ Raise against tPEG
• ☐ Remove WILL DEPEG from the lab to please a deck

6. Suggested build order (so “bigger” has a sequence)

PegLab lab (done: depeg, honesty, KasWare, phone/PC KaChat)
  → live TN10 receipt + indexer
  → KaChat signs it; timeout pay in one thread
  → postage / grams so zero-KAS receivers work
  → groups, tips, stakes, merchant QR, agents
  → mainnet receipt after review
  → hosted USD only if an issuer is on Kaspa
  → overcollateral protocol only with Maker/Liquity-class budget
  → licensed USD only as a company

Skipping to “mainnet chat-pay with a dollar glyph” is how a bridge IOU gets minted.

7. Short version

PegLab is the proof that a cheap pool is not a dollar. Keep it. Deploy something else on mainnet:

• First, a KAS receipt (copy Liquity’s *redemption-is-the-product*, drop the dollar).
• Later, host a named USDC/USDT (copy Sky’s PSM, inherit their risk, name it).
• Only a company copies Circle.

Fund grants and payroll for infrastructure. Fund equity for a bank. Never fund a sister token. Think as big as KaChat-in-the-pocket, merchants, and agents — the covenant is the smallest part.

PegLab stays the lab. The successor earns the mainnet flag.