# Portfolio Rebalance

> Periodic portfolio rebalancing: Optimize → Compare → Execute. Supports MVO, Risk Parity, and Black-Litterman optimization.

- Skill: `knuckles-team/portfolio-rebalance` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add knuckles-team/portfolio-rebalance`
- Raw SKILL.md: https://api.skillmd.com/api/skills/knuckles-team/portfolio-rebalance/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Knuckles-Team (https://skillmd.com/u/knuckles-team)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/knuckles-team/portfolio-rebalance

---

# Portfolio Rebalance Workflow (Cron: Monday 9AM ET)

## Workflow Execution Steps

### Step 1: fetch-positions
Get current portfolio state from exchange.
Tool: `emerald_portfolio(action="positions")` + `emerald_portfolio(action="account")`

### Step 2: optimize
Run portfolio optimization (MVO/Risk-Parity/Black-Litterman).
Tool: Route to `data-science-mcp` for heavy optimization compute.

### Step 3: attribution
Brinson decomposition of recent performance.
Tool: Route to agent-utilities `profit_attribution.py`.

### Step 4: generate-trades
Compute rebalancing trade list from target vs current weights.

### Step 5: risk-check
Pre-trade risk validation on all rebalancing trades.
Tool: `emerald_risk(action="drawdown_check")`

### Step 6: execute
Submit rebalancing orders (paper mode default).
Tool: `emerald_orders(action="submit", ...)` for each trade.

## Execution

Run this workflow as a dependency-ordered DAG. Steps with no unmet `depends_on` run in parallel; dependents run after their prerequisites complete.

- **Run first (in parallel):** Step 1 — fetch-positions; Step 2 — optimize; Step 3 — attribution; Step 4 — generate-trades; Step 5 — risk-check; Step 6 — execute

**Execution:** If graph-os is reachable, offload the whole DAG via `graph_orchestrate action=execute_workflow` (or the `kg-delegate` skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet `depends_on` in parallel, then their dependents.