Are portfolio positions cached or persisted?

No. The final workflow result bypasses Agent result-cache reads, writes, and single-flight, and raw or normalized positions are not placed in persistence, logs, analytics, or public errors.

Can analyze-portfolio-risk recommend a target allocation?

No. It returns current factual calculations and deterministic risk flags without allocation fields, rebalancing instructions, recommendations, or trade execution.

Can AI predict stock prices with this MCP tool?

No. AI cannot predict stock prices with certainty, and this MCP tool does not promise future returns. It gives an agent bounded data or validation results that still require independent review, risk controls, and source checking.

Agent workflows

Portfolio Risk Analysis MCP Tool for Long US Stocks

analyze-portfolio-risk calculates factual risk context for 1–15 positive long positions in active US-listed companies or ADRs. It uses USD prices and costs, reports inclusion coverage, weights, concentration, dividend-adjusted return risk, event facts, and bounded options-risk enrichment, but does not suggest allocations or trades.

Connect Pineify MCP View all MCP tools

MCP tool name

analyze-portfolio-risk

Example prompt

Use Pineify analyze-portfolio-risk for 2 AAPL shares at a $180 average cost and 1 MSFT share at $420. Explain concentration, correlation, volatility, and event flags without suggesting allocations.

Direct answer

What analyze-portfolio-risk does

Each position requires an exact ticker, positive quantity, and positive USD average cost. Duplicate normalized symbols, short positions, options positions, funds, inactive securities, and non-USD instruments are rejected or excluded under the strict contract.

The final portfolio result bypasses the Agent result cache, persistence, logging, and analytics. Provider-owned data caches may still serve public market facts, but raw positions and portfolio-derived keys are not stored.

Contract

What the agent sends and receives

Inputs

`positions`: an array of 1–15 unique exact symbols, each with positive `quantity` and positive `averageCostUsd`.
The contract is long-only and USD-only. It does not accept target weights, options positions, cash balances, currencies, account IDs, or strategy instructions.

Structured output

`portfolio-risk-v1` structured content and `portfolio-risk-methodology-v1` calculation metadata.
Included positions, portfolio totals, weights, sector exposure, HHI/effective positions, historical volatility/correlation facts, and evidence-rich risk flags.
Coverage counts and nullable values distinguish unsupported positions, missing previous closes, insufficient history, zero benchmark variance, and optional options data.

Capabilities

Where this MCP tool fits

Calculate USD portfolio totals

Returns cost basis, market value, unrealized P&L, available day P&L, weights, and included or excluded counts.

Measure concentration and historical risk

Uses HHI, effective positions, sector exposure, 126 calendar-day adjusted returns, volatility, SPY beta, covariance, and bounded pair correlations.

Add bounded event and options context

Enriches included holdings with valuation and earnings facts and requests IV/Greeks/OI context only for the five highest-weight positions.

Agent workflow

A bounded call from question to review

1
Send a minimal long portfolio
Provide only the exact symbols, positive quantities, and USD average costs required for the calculation.
2
Reconcile inclusion and totals
Check included and excluded counts, then confirm returned market values and weights before reading risk flags.
3
Read evidence, not advice
Use the observed values, thresholds, affected symbols, coverage, and source times without converting them into automatic allocation changes.

Prompt examples

Questions an agent can route to this tool

"Use Pineify analyze-portfolio-risk for 2 AAPL shares at a $180 average cost and 1 MSFT share at $420. Explain concentration, correlation, volatility, and event flags without suggesting allocations."

"Analyze this five-stock long USD portfolio. Reconcile the position market values with the total and list every excluded symbol or unavailable risk section."

Operator notes

How I review the result

I send only the position fields needed by the strict schema and avoid placing account identifiers or notes in the request.

I distinguish provider data caching from the final workflow result: the latter does not use result-cache read, write, or single-flight.

Boundaries to keep in the prompt

Only 1–15 positive long positions in active US-listed companies or ADRs with USD market data are supported.
The tool does not support options holdings, shorts, cash, funds, multi-currency portfolios, tax analysis, optimization, target allocations, automatic rebalancing, or execution.
Historical statistics require sufficient paired observations and can be null. They describe the selected lookback and are not forecasts.
This is factual risk analysis, not personalized investment advice.

Pineify MCP is an information and code-validation tool, not investment advice. It does not promise returns, place trades, or modify a portfolio. Review timestamps, source fields, code diagnostics, and risk assumptions before acting.

MCP tool family