Autonomous Lab
Hermes multi-agent research — paper-only
PAPER TRADINGSIMULATED EXECUTIONNO REAL MONEYNOT FINANCIAL ADVICE
Autonomous Paper Portfolio
PAPER TRADINGSIMULATED EXECUTIONNO REAL MONEY
No autonomous portfolio yet — it is created on the first canary campaign.
Portfolio performance — OntoAlpha Autonomous vs SPY vs QQQ (PAPER)
No snapshots yet — the chart fills in after the first daily campaign.
Benchmarks normalized to portfolio inception (—), not raw ETF prices.
Current positions (0)
No open positions.
Autonomous transactions (0) — SIMULATED
No autonomous paper transactions yet.
System
Execution mode: —
Real execution: ENABLED
Campaigns: 0 total, 0 active
Kill switch: off
Hermes runtime
Active runtime: —
Hermes available: no (fails closed)
Feature flags
Campaigns
No campaigns yet.
Decision workbench
Select a campaign to inspect agents, consensus, decisions, and the audit trail.
Shadow portfolios (hypothetical — SIMULATED, never executed)
No shadow portfolios yet.
Compared against SPY · QQQ · equal-weight · no-trade · current strategy. No real money.
Paper canary (isolated ledger — SIMULATED, no real money)
Executes paper trades on a SEPARATE canary portfolio — the active paper portfolio is never touched.
No canary portfolios yet.
Strategy registry (DRAFT → BACKTEST → SHADOW → PAPER_CANARY → PAPER_ACTIVE)
No strategy versions yet.
Promotion is decided by deterministic multi-metric gates — no stage skipping, no agent self-promotion.
Learning center (governed memory — agents propose, evaluator promotes)
No memory candidates yet.
No single-trade promotion. Min sample size + out-of-sample + regime stability + cost-adjusted gates.