The whole project on one page · 00_EXECUTIVE_SUMMARY.md
QALS Project — Executive Summary
Date: 2026-09-06 • Prepared by: qalcode autonomous deep-research session Corpus: 16 cloned IOTA-ecosystem repos (~550MB), live web research (all claims dated/linked in companion docs), 11 design documents.
The one-paragraph answer
IOTA in 2026 is two stories: a failed token (price $0.041, rank #137, all-time-low last month, Shimmer staging network sunsetting 30 Sept 2026) wrapped around excellent, free, Apache-2.0 infrastructure — a Sui-derived Move + EVM object-ledger with provably-live DAG-BFT consensus (Starfish, Apr 2026), a mature identity/trust framework, and real-world trade deployments (TWIN in Kenya customs, ADAPT with AfCFTA/WEF). Our play: fork their node, not their tokenomics. Build Qalnet — qalarc's own private chain across the existing Tailscale fleet — with QALS (4.6B fixed supply, smallest unit "doof") as prepaid service credit, data-anchoring, compute tracking, agent identity, NFTs and an internal DEX; keep everything legally structured as prepaid services until a licensed exchange/stablecoin path is cleared.
Answers to each question asked
| Question | Answer | Doc |
|---|---|---|
| How do we make our own coin/system off IOTA for data transfer? | Fork iotaledger/iota (Apache-2.0, Sui-lineage). Private chain "Qalnet", 4 validators on our own machines, data = anchor objects, integrity anchored to public IOTA. |
03_qals_architecture/ |
| Ternary systems review? | Fascinating history (Curl/Troika/Setun, MIT DCI feud), mathematically elegant, commercially dead — binary everywhere; keep a "trit" branding homage only. | 02_ternary_systems/ |
| All IOTA app components? | Full component map of 17 repos incl. the node, SDKs, wallet, identity, Stronghold, wasp, streams, Firefly — with fork/drop verdicts. | 01_iota_deep_research/IOTA_COMPONENT_MAP.md |
| Credit/token platform across all our apps? | qal_credit Move contracts: CreditAccount + Holds (pre-auth) + subscription streams + on-chain PriceTable; every app and AI agent gets a chain-enforced wallet with spend caps. |
04_credit_token_platform/ |
| DeFi trading interface? | "Qalx": Phase A custom AMM (~1k lines Move) + dApp-kit UI; Phase B port DeepBook v3 (verified Apache-2.0) as CLOB "Qalbook"; treasury-seeded liquidity, burn-on-usage. | 05_defi_trading/ |
| Zip-style bank/exchange? | 3-phase: closed-loop prepaid credit (no licence) → AUSTRAC DCE registration + qAUD stablecoin with partner → ACL credit (Zip-style instalments) / e-money path. Full cost/licensing reality check inside. | 06_bank_exchange/ |
| Can a tangle coin transfer & track compute? | Yes — compute becomes first-class objects: escrowed ComputeJob lifecycle, provider attestations, benchmark oracle, slashable stake, tradeable GPU-hour vouchers, receipt NFTs, AI-output provenance. | 07_compute_marketplace/ |
| NFT structures + our NFT projects? | IRC27 (legacy) → Move objects + Kiosk/TransferPolicy/Display + Gas Station gasless mints (current). Ship compute-receipt NFTs first, Doofs PFP later. | 08_nft_strategy/ |
| Digital identity + AI identity? | "Qal ID": DIDs for users AND agents, capability VCs, RFC 9421/Web Bot Auth for agent auth, on-chain revocation (~2s kill switch), SD-JWT selective disclosure, Hierarchies-based issuer governance. | 09_identity_ai/ |
| Current state + news of IOTA? | Rebased mainnet 5 May 2025 (Move L1, DPoS, coordinator gone); supply 4.6B but cap abolished (inflation); Starfish consensus Apr 2026; trade-infra pivot (TWIN/ADAPT/WEF/GLEIF); token at ATL. Shimmer sunsets 30 Sept 2026. | 01_iota_deep_research/STATE_OF_IOTA_2026.md |
The five decisions that matter
- Private fork, not public token. IOTA's decade proves public-token economics destroy good infrastructure. QALS = prepaid credit on a chain we run; sovereignty without speculation.
- Fixed 4.6B supply, no protocol inflation. (IOTA abolished its cap — we won't. Validator rewards come from a finite allocation.)
- Move for value, EVM deferred. Move VM from day one; the EVM surface is a separate IOTA stack (not in the base repo) — add later via the EVM-as-core-contract pattern if Solidity compatibility earns its keep.
- The AU$1 backing is structural, not promised. Two token classes: B-QALS minted only against AU$1 in a segregated reserve (redeemable at AU$0.995, burned when consumed — gift-card economics on-chain) and G-QALS (growth allocation under a revenue-funded ratchet floor). Full design:
03_qals_architecture/QALS_BACKING_DESIGN.md. - Agents are first-class users. DID + wallet + spend-cap + revocable credentials — chain-enforced limits beat prompt-engineered ones.
- Compliance gates the public surfaces. Phase-1 B-QALS are non-transferable prepaid credit (no licence path required per doc 06 analysis); nothing publicly tradeable until the AUSTRAC/VASP path is executed.
Costs & timeline (headline)
- Weeks 1–4: devnet on superlocal, fork renamed, QALS genesis, first DataAnchor from the agent hub. ~$0 (labour only).
- Months 2–3: The Fleet validators + core contracts + hub metering live. ~$50–100/mo cloud.
- Months 4–6: all apps integrated, compute settlement, ComputeReceipt NFTs, Qalx Phase A.
- Month 6+: AUSTRAC DCE + qAUD (per doc 06), Doofnet public testnet, DeepBook port if demand exists.
The honest risk list (short version)
Regulatory reclassification of QALS (top risk — mitigated by prepaid framing + legal review before public sale); fork-maintenance burden (mitigated: our changes live in qal_* Move packages, core stays upstream); home-machine validators (mitigated: Starfish lag-tolerance + cloud validators); Move/Rust skill ramp (mitigated: IOTA workshops/docs are Sui-class, EVM fallback exists). Full table in doc 03 §7.
Index: 01 IOTA research • 02 ternary • 03 QALS architecture (+ AU$1 backing design) • 04 credit platform • 05 DeFi (+ Qalx platform) • 06 bank/exchange • 07 compute • 08 NFTs • 09 identity/AI • 10 roadmap. Deliverables: wiki/ (full site) • dist/ (PDF suite) • interactive/qals_explorer.html (interactive diagrams) • repos/ (16 cloned IOTA repos) • data/ (GitHub org metadata) • scripts/ (clone + build tooling — preserved generation code).