Onboarding checklist

1. Portfolio foundation (non-negotiable)

Most recommendation accuracy comes from the book the product can see: portfolios, accounts, positions, and cash.

• Create or designate primary portfolio(s). You can run multiple books (for example Core Income, Growth Alpha, taxable vs IRA). The workspace tracks each independently with lot-level cost basis.
• Import or connect every account. Broker CSV upload (Merrill, Fidelity, etc.) via /import-activity — auto-mapping builds consolidated holdings with Avg cost vs Last.
• IBKR Client Portal (recommended for a live execution path): Connect IBKR — consent plus sealed session surfaces positions, cash, and margin for entitled accounts.
• Verify consolidated holdings in Portfolio→ Edit Account: cash balances, open positions, average costs, and day P&L.

Covered calls, cash-secured puts, collars, and wheel-style workflows only line up with your capital and underlying exposure when the engine sees real positions. Without that, even xAI stays theoretical.

2. Curate a high-conviction watchlist

Watchlist drives symbol-scoped discovery and feeds downstream scanners and xOptions context.

• Open /watchlist and add: all core holdings; 10–20 high-conviction names you trade or want flow on; names tied to your outlook (earnings, event-driven ideas, etc.).
• Enable price alerts on key levels — scheduled scanners and desk alerts surface opportunities when configured.
• The workspace rail compact watchlist and "hot" IV/OI rows feed xOptions Step 1 (Find Options) and multi-agent workspace context.

3. Risk profile and investment outlook

Explicit posture maps into conservative / balanced / aggressive scoring paths and strategy jobs.

• In xChat (or your initial NL prompt), state: risk tolerance (conservative income / defined-risk, balanced theta + mild delta, aggressive directional or vol-selling — only if your role and broker approvals allow); market outlook (bullish, range-bound, bearish, event-driven, vol crush, etc.).
• Add time horizon, allocation rules (e.g. max notional per strategy, preferred DTE bands), tax and liquidity constraints where they matter.
• Seeded options strategy preferences (Mongo catalog aligned with disk xPersonas / coreskills) plus your chosen xChat persona anchor those signals into strategy scoring and StrategyJob-class flows — preflight → slot collection → synthesize becomes surgical instead of generic when workspace data is complete.

4. Broker and market-data parity

Keep custodian identity and imports aligned so sync and desk views match the broker of record.

• Update the broker / external account reference (ext account id and related desk fields) so each portfolio account syncs to the right custodian row.
• Re-run broker importwhen you need the latest CSV-backed activity; use consolidated holdings and import previews to reconcile vs the broker's export.
• Where IBKR is linked, refresh consent/session if snapshots look stale so positions and cash match Client Portal.

5. Personalize xChat and xOptions

Continuity and product-specific personas tighten recommendations across sessions.

• In xChat preferences, opt in to conversation history when available — turns persist in Mongo xchat_logs with a 60-day TTL pattern on stored turns (product-enforced rolling window).
• Use the options-strategy persona (or the hardcore strategy path) for structure-heavy asks so routing uses RAG + OptionsStrategyEngine scoring where configured.
• On /xoptions, run at least one full stepped builder flow — Step 1 bootstrap pulls portfolio + watchlist context and seeds context for later jobs.

6. Final validation and “Run now” test

Prove the stack end-to-end: scheduler or strategy job, then inspect outputs like you would a desk review.

• Trigger execution: Have an operator run a manual scheduler tick where your tenant uses scheduled jobs (Admin → Tasks for global_admin; workspace automations at /workspace/tasks when configured). Or start a new StrategyJob from xChat — for example: Run full options strategy scan on my Core Income portfolio using balanced outlook with 2-week horizon.
• Review outputs: strategy cards, payoff charts (Apex in-product), Greeks, and risk / outlook tags — confirm they match your book and the posture you stated in step 3.
• If something looks off: the admin audit trail (Admin → Audit when you have access) plus tenant xChat debug logging (Admin → Tenant workspace, when enabled) make iteration and refinement straightforward for operators — no guesswork.

Expected outcome once complete

With the steps above done, you should see production-grade, xAI-augmented recommendations that:

• Respect your actual portfolio construction and cash — grounded on consolidated holdings, accounts, and imports you verified.
• Align with your chosen risk bucket — conservative, balanced, or aggressive — as you expressed in xChat and scoring prefs.
• Leverage the full stack — RAG-linked options coreskills library plus OptionsStrategyEngine scoring where the product routes your asks.

Outcomes still depend on market data freshness, persona/RAG configuration, and plan limits — treat this as the standard setup bar, not a performance guarantee.