Checking...
Components
API Server
...REST endpoints that back the terminal, vault, and agent flows
WebSocket
...Real-time signal feed, position updates, vault events
Solana RPC
...Live chain reads for balances, slots, and transaction confirmation
Jupiter Aggregator
...Quotes and swap routes for every trade the agent and users execute
Redis
...In-memory cache for hot data and inter-process pub/sub
Vault Reconciliation
...Drift detector running every 5 minutes against on-chain state
Telegram Bots
...Public Pulse channel posts and operator alert delivery
Severity Reference
Responding within expected latency, no anomalies detected.
Functional but slower than baseline, or partially impaired. No action required if state persists briefly.
Component unreachable or failing health checks. Trade flows that depend on this component may be affected.
Optional component intentionally disabled, e.g. Telegram bots when channel id is unset. Does not affect overall status.
What happens during an outage
- API or WebSocket down — the terminal will show stale data. Already-open positions continue to be managed by the server-side position monitor; new trades cannot be entered from the UI until the API recovers.
- Solana RPC degraded — quotes, balances, and confirmations slow down. The vault and Solo Agent throttle new entries while RPC is slow.
- Jupiter unreachable — no new swap routes can be computed. Open positions still exit via stop-loss and timeout because exit routes are computed at trigger time, not in advance.
- Vault reconciliation red — the vault is auto-paused. No new deposits or auto-trades. Withdrawals of idle SOL continue. An admin must investigate before resuming.
- Redis down — caches refill from source on first miss. Hot endpoints get slower until Redis is back, but no data is lost.
- Telegram down or not configured — public Pulse channel posts and operator alerts are not delivered. The server logs them either way, so nothing is lost from a forensic standpoint.
Verify yourself.
This page reads from the same status API that external uptime monitors poll. Point your own monitor at /api/status if you want a second opinion.