PowerChain Renewable RWA + Staking v1.0.0 is a production-oriented, multi-chain platform for PWRC staking, renewable-infrastructure participation, rewards, governance, network assurance, portfolio operations, payments, and evidence-grounded AI assistance.
Status:
v1.0.0·devnet+ gatedmainnet-beta· execution fail-closed by default
PowerChain ships as one monorepo containing the public web experience, authenticated dashboard, backend/API, workers/indexers, shared packages, Solana programs, Sui Move package, accounting/reconciliation services, and secured Copilot runtime.
| Domain | Core capabilities |
|---|---|
| Staking | Flexible/fixed PWRC staking, increase, claim, compound, unstake, cooldown/unbonding, withdraw, indexed positions |
| Rewards | Funded epochs, cumulative indexes, vault solvency, claim accounting, immutable reward receipts |
| Governance | Proposals, voting weight, signed vote intents, quorum/timelock policy, reconciled execution |
| Network | Operator/verifier bonds, slashable service collateral, oracle/DePIN visibility, runtime health/readiness |
| PWRC | Solana Token-2022 validation, balances, transfers, Solana Pay, trusted-token policy |
| Cross-chain | PWRC ↔ wPWRC reserve verification, Sui gRPC, Sui Move bridge, CCTP V2 boundary |
| Portfolio | Accounts, wallets, tokens, balances, positions, activity, histories, fees, claims |
| Liquidity | Jupiter, Meteora, Orca, Raydium, Cetus and DeepBook integration boundaries |
| Market data | Pyth, CoinGecko, CoinMarketCap, Birdeye and explorer/provider adapters |
| Identity | Clerk primary auth, optional Better Auth, social login, wallet ownership signatures, account-wallet binding |
| Security | Exact-action confirmations, replay protection, rate limits, runtime kill switches, audit/security events |
| Accounting | PostgreSQL, Prisma, double-entry ledger, reconciliation exceptions, hash-linked immutable receipts |
| AI Copilot | Multi-provider LLMs, RAG, MCP read tools, local/LoRA, evals, grounding/citation enforcement, abstention |
| Operations | Helius/webhook ingestion, workers, indexers, observability, circuit breakers, recovery/admin controls |
PowerChain
├── apps/
│ ├── web/ public product, Learn, legal
│ ├── dashboard/ authenticated operational application
│ └── backend/ API, chain services, accounting, workers, AI
├── packages/
│ ├── ui/ canonical UI tokens and primitives
│ ├── contracts/ shared API/runtime contracts
│ ├── api-client/ typed client boundary
│ ├── core/ domain utilities
│ └── integrations/ provider abstractions
├── programs/
│ ├── solana/ staking, governance, bonds, bridge
│ └── sui/ wPWRC Move package
├── skills/ AI assistants, agents, skills, character policy
├── docs/ architecture, API, security, operations, production
└── scripts/ release/security/AI/design/integration audits
The dashboard proxies same-origin /api/* requests to the backend. Provider credentials, signing policy, RPC secrets, database credentials and service-role keys remain server-side.
PowerChain separates account identity, wallet ownership, transaction authorization, accounting truth, and AI authority.
Clerk / Better Auth
↓
wallet ownership challenge
↓
account ↔ wallet binding
↓
server-authoritative intent
↓
exact-action confirmation when required
↓
prepared transaction integrity
↓
user wallet signature
↓
chain finality
↓
reconciliation
↓
double-entry ledger
↓
immutable receipt
Social login authenticates the account only. It never grants wallet spending authority. The server does not custody user wallet private keys.
AI/MCP is read-oriented and independently policy-gated. AI cannot access MPC/threshold key shares, sign wallet messages or transactions, stake, claim, transfer, bridge, vote, or bypass readiness/reconciliation/accounting controls.
Overview Copilot Stake Positions
Rewards Claim PWRC Governance
Network Portfolio Accounts Wallets
Histories Activity Leaderboard Integrations
Bridge Explorer Profiles Settings
Security Learn
Account-specific balances, positions, rewards, votes and transaction states must come from authoritative backend/indexer data. Representative aggregate data must be explicitly labelled.
PWRC Token-2022 · Anchor staking/governance/bonds/bridge · Helius/configured RPC · Solana Pay · Jupiter · Meteora · Orca · Raydium · Pyth · Metaplex/DAS boundary · Solscan · Circle CCTP V2
wPWRC · Sui gRPC · Move bridge package · Cetus · DeepBook · Suiscan
PostgreSQL · Prisma · Supabase boundary · Clerk · Better Auth · CoinGecko · CoinMarketCap · Birdeye
All integrations are server-side policy boundaries. Missing production credentials or unsupported capabilities must fail closed rather than fabricate state.
Canonical namespace:
/api/v1/
Primary domains:
auth system staking positions
actions rewards claim governance
bonds wallet transfers portfolio
accounts histories prices swaps
solana-pay cctp bridge programs
integrations tokens ai admin
Service health and execution readiness are deliberately separate:
GET /api/v1/system/health
GET /api/v1/system/readiness
A healthy service does not automatically mean state-changing execution is permitted.
See API.
PowerChain uses a dedicated authenticated WebSocket process for fast lifecycle invalidation while keeping HTTP, PostgreSQL, reconciliation and the double-entry ledger authoritative.
PostgreSQL transaction
→ minimal realtime event
→ NOTIFY/LISTEN
→ authenticated WebSocket
→ dashboard invalidates query
→ GET /api/v1/... authoritative refresh
Bootstrap and discovery:
GET /api/v1/endpoints
GET /api/v1/realtime/status
POST /api/v1/realtime/token
GET /api/v1/realtime/events
Run the transport independently:
npm run dev:realtime
npm run realtime:start
npm run realtime:checkThe socket carries status/invalidation metadata only. It never becomes the source of truth for balances, positions, rewards, governance weight, bridge coverage, receipts or settlement. See Realtime.
PowerChain Copilot uses an evidence-first pipeline:
request
↓
identity + AI policy
↓
trusted RAG + verified read tools
↓
evidence scoring + typed tool contracts
↓
allowlisted model/provider
↓
DLP + grounding/citation verification
↓
VERIFIED / GROUNDED / INFERENCE / UNAVAILABLE
The runtime includes provider health/circuit breakers, failover, immutable prompt fingerprints, trust-tiered RAG, MCP read-only tools, local/LoRA adapters, bounded memory, output DLP, policy evals, progress streaming and explicit abstention when evidence is insufficient.
No LLM can guarantee literal zero hallucinations; PowerChain instead prevents unsupported operational claims from being presented as verified facts.
See AI Runtime, AI Evals, and AI Security.
State-changing execution is fail-closed:
EXECUTION_ENABLED
↓
valid deployment configuration
↓
runtime control enabled
↓
fresh readiness attestation
↓
PWRC Token-2022 valid
↓
programs verified
↓
reward vault solvent
↓
PWRC/wPWRC reserve covered
↓
workers/indexers healthy
↓
no blocking operational alert
↓
EXECUTE
Critical configuration changes invalidate previous readiness evidence. Runtime controls can pause staking, governance, bonds, transfers, rewards, bridge, AI, or global execution without redeployment.
PowerChain uses a restrained monochromatic system with controlled operational green:
Canvas #ffffff
Soft surface #f5f5f5
Subtle surface #fafafa
Primary #0a0a0a
Secondary text #525252
Muted text #737373
Border #e5e5e5
Operational green #0b6b38
Destructive #e7000b
Black is the default primary action. Dark green is reserved for PowerChain operational/verified/energy semantics. Red is destructive/failure only. Interactive geometry is 18px; cards/containers are 24px. No gradients, glassmorphism, colored shadows or decorative chromatic palettes.
See Design System and UI/UX.
- Node.js
>=22 - npm
10.9.x - PostgreSQL
- Anchor/Solana CLI only when building Solana programs
- Sui CLI only when building/testing Move programs
npm install
cp apps/backend/.env.example apps/backend/.env.local
cp apps/dashboard/.env.example apps/dashboard/.env.local
cp apps/web/.env.example apps/web/.env.local
npm run prisma:generate
npm run db:migrate
npm run ai:ingest
npm run devRun applications individually with:
npm run dev:web
npm run dev:dashboard
npm run dev:backendSupported deployment profiles:
devnet
mainnet-beta
Mainnet-beta remains disabled until real PWRC/program/vault/bridge values, RPC/provider credentials, auth secrets and service integrations are configured and verified. Never promote placeholder program IDs, sample wallet addresses or representative data into production.
See Production and Programs.
PostgreSQL is the authoritative application/accounting store. Finalized chain transactions are reconciliation evidence, not a replacement for the internal ledger.
npm run worker:reconcile
npm run worker:reconcile:loop
npm run worker:index
npm run worker:index:loopThe persistence layer includes staking/reward/governance/bond state, identity/wallet bindings, double-entry journals, immutable receipts, reconciliation exceptions, webhook observations, provider telemetry, AI/RAG state and worker/indexer health.
Dependency-free platform validation:
npm run validate:releaseThis runs source, program compatibility, integrations, auth/security, AI runtime, canonical design-system, UI integrity and platform-release audits.
Production release additionally requires:
npm install
npm run prisma:generate
npm run prisma:validate
npm run typecheck
npm run build
npm run programs:build
npm run programs:test
npm run programs:sui:build
npm run programs:sui:testDo not set EXECUTION_ENABLED=true until /api/v1/system/health and /api/v1/system/readiness are healthy and the intended deployment has verified Token-2022 configuration, executable programs, reward-vault solvency, bridge reserve coverage and worker freshness.
| Document | Purpose |
|---|---|
| Architecture | System boundaries and service topology |
| API | Backend/API contracts |
| Realtime | WebSocket protocol, invalidation model and deployment |
| Production | Deployment and mainnet-beta gates |
| Operations | Runtime operations and recovery |
| Observability | Tracing, metrics and operational telemetry |
| Security | Platform security policy |
| Auth Security | Identity, wallet and confirmation model |
| Programs | Solana/Sui protocol workspace |
| AI Runtime | Copilot/RAG/MCP/provider architecture |
| AI Security | AI authority and safety boundaries |
| Design System | Canonical UI tokens and rules |
| Release Validation | Verified release gates and limitations |
Canonical product version: 1.0.0.
Production hardening, protocol wiring, AI-runtime improvements and UI refinement continue under the v1.0.0 baseline unless a product-level version change is explicitly approved. Release history belongs in CHANGELOG.md; this README remains the concise product and engineering entry point.