System Settings

Saved

Trading Mode

DRY
LIVE places real orders on funded exchanges — irreversible. Per-binding run_mode overrides it: auto inherits this switch; live/dry are pinned.
Fleet run-mode
auto live dry stopped
enabled bindings pinned — unaffected by the global switch until set to auto.
Switching to LIVE activates every auto binding (read-time, no DB write). Pin individual bindings to dry in Smart Funding / Pairs, then Save SettingsApply to Engine.

Inference Configuration

Tuning Server — inventory pull

The Market page pulls its shard pool, tradeable symbols and timeframes from here. Nothing is hand-entered — you add a (symbol · timeframe) to an exchange from this pulled pool.

Save first, then test.

Node fire-entry feed — inbound push

The tuning node POSTs gate-passed fire entries to /api/fire-entries on this client. This shared secret authenticates that push — set the same value on the tuning node's push config.

AIServer L2 — hold / exit gateway

Each cycle the client sends open-position state to the AIServer SAC gateway and applies the returned HOLD / EXIT. Usually loopback when co-hosted with AIServer.

Engine Status

Status-
Mode-
Cycles-
Last Cycle-
Exchanges-

Gateway Status

auto-refresh 5s

L1 Backends

No gateway configured

L2 Backends

No gateway configured

GitHub Sync

checking…
Auto-sync pushes only — it never pulls, so a remote edit can't overwrite this machine. Use Pull manually when another box has changes (fast-forward only — aborts on divergence, nothing overwritten); push is never forced. Secrets, data, logs and model weights are excluded by .gitignore.