What each number means.

• TVL — sum of liquidity-well balances across all live wells, read via /api/onchain-tvl. This route reads each LWC's stable balance on-chain directly, no aggregator.
• Active networks — count of wells reporting non-zero balance and a recent heartbeat. Pulls from the same TVL route.
• 24h volume / median fill — from the fill ledger (arc-api /v1/fills). Aggregated over a rolling 24h window.

If a number says "data unavailable", the underlying route returned an error or an empty payload. The portal withholds totals instead of inventing a number — by design.

The orders list tails the genome stream and shows your intents by source/dst chain, asset, and stage:

• · pending — intent submitted, in bid window
• · filling — solver selected, waiting for dst-chain inclusion
• · settled — destination tx confirmed, settlement committed
• · refunded — no solver filled within the refund window, source-side balance returned

Each row links to the per-order page /bridge/orders/[id] which shows the full tx graph (source tx, fill tx, settlement tx, refund tx if any).

• /explorer/orders — every settled fill, paginated. Source for the live stream component on /stats.
• /explorer/executors — solver leaderboard. Ranked by 24h / 7d / 30d window. Click a row for per-solver fill history and median latency.
• /explorer/protocols — adapter health: live vs dormant, recent fills, fee bps. Source: solver_intel.json.

Both panels read directly from the contract — no API in the critical path:

• Your TRN, govTRN, allowance, pending rewards, total staked — five on-chain reads via wagmi useReadContract.
• Symbiotic shares + assets — vault shares are 1:1 with deposit until withdrawals start. Active stake is the bonded portion currently securing solvers.

If a stake panel says "GovStakingV2 is not configured for the current environment", the env-mode switch in the header is set to a mode without a deployment. Today only prod (Arbitrum) and devnet (Taifoon) are configured.

The arc-api dashboard exposes operator views atlocalhost:8082/dashboard (or via the public reverse proxy on the arc host). It shows:

• per-chain RPC health
• fill-monitor lag (how far behind real-time the scanner is)
• deployer queue depth + last successful deploy
• donut accrual per registered adapter
• recent Telegram notification history (settled fills, large transfers)

1. Cross-check on Arbiscan / Etherscan. Every contract address used by the portal is published in contracts/deployments/*.json and in the per-section docs.
2. Check /api/onchain-tvl directly in your browser. If the route returns 5xx, the issue is server-side (RPC degraded). If it returns 200 with sensible numbers, the issue is in the rendering layer.
3. Check arc-api /v1/health for per-chain RPC liveness.
4. Open a Discord ticket with the route, the rendered value, and the on-chain value. The most likely cause is an aggregator cache stale by < 60s.