LUME Protocol — Development Report (June 2026)
TL;DR
We took the LUME concept from discussion into a working CKB testnet prototype (v0.1.1). Users can connect wallets, stake CKB via the official iCKB / NervosDAO path, accrue simulated LUME yield, and claim LUME as RGB++ testnet xUDT in real on-chain transactions. BTC entry is architected as Lightning ↔ Fiber atomic swaps (not RGB++ BTC mapping), with Omnilock identity and Fiber/PTLC ingestion intents in place. Rust contract logic for subsidy shards, state rent, credit markets, and four systemic hardening modules is implemented and unit-tested — deployment to CKB testnet is the next milestone.
You can Test the Demo: https://lume-protocol.vercel.app/ , it is all still under experiment
1. Alignment with the Thread Vision
The original thread framed LUME as a programmable yield layer bridging Bitcoin and CKB via RGB++, iCKB, and Fiber. Our implementation follows that direction with one important technical correction raised by @Ckroamer:
RGB++ binding is CKB → BTC, not native BTC → CKB.
So BTC participation is routed through trustless cross-chain atomic swaps between Bitcoin Lightning and CKB Fiber — . That is the foundation of our BTC entry path.
| Thread concept | LUME prototype status |
|---|---|
| Nervos DAO / iCKB synergy | Live — real stake txs via @ickb/sdk |
| RGB++ LUME token | Live — testnet xUDT deploy + mint |
| 21M supply cap | Enforced in claim flow |
| 10% APY floor (3 layers) | Modeled — base + active + subsidy; accrual simulated client-side |
| Fiber Network routing | Architected — ingestion intents + withdrawal fee curve; not wired to a live Fiber node yet |
| Lightning ↔ Fiber swaps | Preview — message-signed atomic swap intents |
| BTC collateral / Omnilock | Partial — Omnilock BTC wallet, collateral monitoring, JoyID/CKB path separated |
| AI agentic hooks | Roadmap — tiered hook design discussed in thread; not implemented |
| Miner security dividends | Roadmap — keeper incentive math stubbed |
| Hardware-verified RWA collateral | Roadmap — open to collaboration (e.g. Luxvoid) |
2. What Ships Today (Testnet Prototype)
Live on CKB testnet
-
Capital ingestion (CKB L1)
- JoyID wallet connect
- CKB → NervosDAO deposit via official iCKB protocol
- Liquid iCKB xUDT received on-chain
-
LUME redemption
- RGB++ testnet xUDT mint to connected wallet
- 21,000,000 LUME hard cap enforced per workspace
-
Operational studio
- Dashboard: stake, yield matrix, capacity rent pool, cell telemetry, activity log
- Wallet-scoped persistence (per CKB address)
- HTTP CKB RPC proxy (
/api/ckb) for reliable browser access
Preview / intent-only (not yet on-chain settled)
- BTC stake —
Bitcoin L1 → Fiber pre-confirmation → LUME ingestion(Omnilock message sign) - Lightning ↔ Fiber atomic swap — interoperability intent
- Mobile money escrow — TZS preview flow
- LUME yield accrual — client-side simulation with offline catch-up (not chain-verified settlement)
Dual-wallet architecture (Option A)
| Wallet | Role |
|---|---|
| JoyID | CKB stake, iCKB, LUME RGB++ claim |
| Bitcoin Omnilock | BTC ingestion stake, BTC yield claim (same wallet required when BTC is staked) |
This mirrors the thread’s goal of sovereign keys without custodial CKB accounts.
3. Protocol Entry Architecture
LUME is designed for multi-path entry:
CKB ──► L1 (iCKB / NervosDAO) ──────────────────────► LUME yield ✅ LIVE
BTC ──► L1 lock ──► Fiber PTLC (L2) ────────────────► LUME ingestion 🔜 planned
BTC ──► Lightning (L2) ──► atomic swap ──► Fiber/LUME 🔜 preview
CKB ──► Fiber channels (L2) ────────────────────────► fast routing 🔜 planned
As @Ckroamer noted, a well-functioning BTC ↔ CKB swap via Fiber is the critical path for the BTC community bridge. That is our top integration priority after testnet contract deployment.
4. Tri-Layer Yield Model (Mapped to Code)
The thread’s Tri-Layer Yield Architecture maps to our three APY layers:
| Layer | Thread description | LUME implementation |
|---|---|---|
| Base | Nervos DAO / iCKB staking | 2.14% share of 10% floor — live iCKB stake path |
| Active | Fiber liquidity / routing premium | 6.48% share — credited on BTC stake + swap intents |
| Subsidy | Algorithmic shortfall top-up | 1.38% — max(0, 10% − (base + active)) via 256-shard matrix |
Subsidy formula (on-chain mirror in Rust):
shortfall = max(0, yield_target − (yield_base + yield_active))
shard_index = last_byte(lock_hash) % 256
Reverse value capture: 70% excess LUME burn / 30% CKB state absorption vault (constants in protocol config; vault not deployed yet).
5. On-Chain Contract Work (Rust / CKB-VM)
Eight Rust modules, 21 unit tests passing:
| Module | Purpose |
|---|---|
subsidy_shard.rs |
256-shard subsidy matrix, 70/30 excess split, 5% emergency gas vault allocation |
state_renting.rs |
300-epoch grace, keeper eviction, SMT compression, commit–reveal top-up branches |
credit_market.rs |
Collateral ratio (150% min), liquidation status bytes, sliding keeper incentives |
smt_archive.rs |
Dynamic SMT ring buffer (10k→50k slots), epoch checkpoint proofs |
fiber_withdrawal.rs |
Anti–bank-run fee curve: Fee = Base × (Threshold/R_pool)³, async OTX fallback |
keeper_gas_vault.rs |
Emergency gas subsidy vault, shard-isolated OTX injection |
commit_reveal.rs |
MEV-blinded cell top-up: Hash(Cell_ID + Salt + Amount) |
Status: Algorithmic logic complete; not yet cross-compiled or deployed to CKB testnet (riscv64imac-unknown-none-elf --features script).
6. BTC Path — Practical Design Choices
Responding to thread questions and community feedback:
- No fake “RGB++ native BTC” — we use Lightning ↔ Fiber atomic swaps for BTC→CKB liquidity.
- Omnilock-only BTC stake — prevents wallet mismatch on BTC yield claims.
- BTC collateral enforcement — wallet L1 balance (Taproot + Native SegWit) monitored against staked amount; yield pauses if undercollateralized until top-up or liquidation intent.
- Fiber/PTLC route documented in wallet intents as:
Bitcoin L1 → Fiber pre-confirmation → LUME ingestion vault
7. Repository & Milestones
| Item | Detail |
|---|---|
| Repo | github.com/TOKENPESA/lume-protocol |
| Tag | milestone-v0.1.1 |
| Network | CKB testnet (Pudge / ckt1) |
| Docs | docs/LUME-Protocol-Documentation.md + PDF |
| Build | cargo test  · npm run build |
Following @toastmanAu’s advice — we built the testnet prototype.
8. Honest Limitations (June 2026)
We prefer transparency over hype:
- LUME yield is simulated, not settled by on-chain vault contracts
- BTC stake and Fiber/Lightning paths are message-signed intents, not live channel settlements
- Rust contracts are stubs with real math — not deployed type scripts
- AI agentic hooks, miner dividends, RWA hardware collateral — roadmap
- iCKB burn → unlock CKB not yet in-app
9. Summary
LUME is moving from AI-generated idea to testnet operational . The core thesis of the thread — iCKB staking multiplier + RGB++ LUME + Fiber/Lightning BTC bridge — is reflected in working code, with CKB L1 and RGB++ claim paths live today. The hardest and most valuable next step, as the community correctly identified, is a real BTC ↔ CKB swap rail via Fiber — that is where we are focused next.
The Project is Experimental; A lot of things might change in the coming developments