AdvisorAssist
ChatGPTTool for handling queries related to the Direct Advisory Suite. Supports: - Investment Plan and Proposal creation - Client and Group management activities - Portfolio management and related functionalities Tool invocation rules (strict): - Do NOT call other tools or trigger any actions automatically. - Wait explicitly for the user to provide a clear request before proceeding. - NEVER assume the next step or preemptively create investment proposals or queries. - If the tool receives no user command, it must remain idle and refuse to proceed. - Any automatic or unsolicited calls must be flagged as a violation and halted immediately. Parameters: request: QueryRequest object containing the user query string and session_id query (str): The user query to be explicitly processed session_id (str): The session ID for the conversation thread. Pass empty string "" on the first call. On subsequent calls, pass the session_id returned from the previous response to maintain conversation context. NEVER ask the user for a session_id - always manage it automatically. Workflows: - To create a portfolio, provide holdings as a list in otherDetails (each holding has keys: 'identifier', 'marketvalue', 'quantity', 'holding_name'). identifier accepts ticker, security name, CUSIP, or ISIN — ID resolution is handled internally. Guidelines for responses: - Use concise and direct natural language strictly responding to the user query. - Do NOT add unsolicited suggestions or initiate tasks without user command. - Keep responses brief (1 to 3 sentences max). Formatting guidelines: - Use Markdown for all responses. - Use headings (##) for major sections (e.g., "Summary", "Next Steps"). - Use bullet points for lists of items. - Use tables for structured portfolio data. - Use fenced code blocks (``json ... ``) for JSON or technical outputs. - Use blockquotes for callouts: > ⚠️ Warnings > ✅ Success > ℹ️ Information
morningstar-analyst-research-tool
ChatGPTThe Morningstar Analyst Research Tool enables users to access detailed, forward-looking research reports on individual stocks, open-end mutual funds, and ETFs. For stocks, the tool provides comprehensive analysis covering business strategy and outlook, Bulls Say/Bears Say perspectives, economic moat analysis, fair value assessment, risk and uncertainty evaluation, capital allocation analysis, and recent analyst notes and updates. For funds and ETFs, it delivers Morningstar's Medalist Pillar assessments evaluating investment strategy (Process), manager quality (People), asset manager strength (Parent), historical performance, and expense structure (Price). The tool supports research for major domiciles including the USA, Canada, Japan, Australia, France, China, Luxembourg, and many more, but is limited to the latest published report. Use this tool when you need comprehensive and up-to-date Morningstar analyst analysis on specific stocks, open-end funds, or ETFs.
morningstar-articles-tool
ChatGPTUse this tool to answer questions about the latest general financial market news, investing themes, personal finance, and sector or asset class comparisons based on articles written by Morningstar’s editorial team. Choose this tool when you need commentary, trends, or analysis on broad markets, events, investment strategies, official Morningstar methodology, or educational content on investment concepts, specifically as covered by Morningstar from 2022 onwards. This tool is best when you need general financial explanations directly derived from Morningstars official editorial perspective. This tool provides research at a broad level and is not suitable for retrieving specific data points or detailed investment-level analysis.
morningstar-data-tool
ChatGPTThis tool is best used when you need to retrieve specific, structured quantitative metrics for individual securities, such as stocks or funds, using Morningstar’s data. It requires a valid investment ID and a datapoint ID, which can be obtained using only the Morningstar ID Lookup Tool. The tool supports querying for a growing list of datapoints (currently over ~500 datapoints), and can also provide data for select historical periods where available. Use this tool when you need precise investment-level values for a security, but note that the historical data coverage varies depending on the datapoint. If providing a date range for historical data, both start_date and end_date must be provided.
morningstar-data-visualization-tool
ChatGPTUse this tool to render a line chart whenever the user asks to graph, chart, plot, or visualize time-series data. It accepts data as a list of row objects and returns an interactive chart.
morningstar-fund-holdings-tool
ChatGPTUse this tool to retrieve the detailed portfolio composition of a managed investment, specifically their largest holdings and the percentage allocation to each holding. Always call this tool after morningstar-id-lookup-tool when the user asks what a fund holds, what's inside it, or to break it down — the ID lookup alone does not return holdings data. The tool allows you to specify how many top holdings to return (between 10 and 100) and can process requests for multiple funds at once, as long as you have the relevant Morningstar investment IDs. This tool is suitable for analyzing fund transparency, portfolio concentration, and understanding how assets are distributed among underlying securities. Only applicable for Exchange-Traded Funds (FE), Open End Funds (FO), Closed End Funds (FC), and Collective Investment Trusts (CZ).
morningstar-id-lookup-tool
ChatGPTUse the Morningstar ID Lookup Tool to map investment tickers, security names, ISINs, or datapoint names to their corresponding Morningstar internal IDs. This tool is the primary way to convert investment and datapoint information into standardized identifiers required for other tools to work properly. The input to the tool must be the extracted investment tickers, investment names, or ISINs, datapoint names from the user's query. Provide at least one of: investment_identifiers or datapoints. Currently the tool only supports: Stocks (ST), Exchange-Traded Funds (FE), Open End Funds (FO), Closed End Funds (FC), and Collective Investment Trusts (CZ). The investment_type field in results indicates the matched security type. If results do not match the expected type, retry with a more specific or differently worded identifier.
morningstar-portfolio-analysis-tool
ChatGPTUse this tool to analyze a portfolio of holdings or a fund using Morningstar X-Ray API. Accepts either a custom portfolio (stocks, ETFs, funds, or cash) with weights summing to 100%, or a single fund ID to analyze a fund's underlying holdings. Use the analysis_type parameter to select a specific slice of the analysis, such as asset allocation, equity characteristics, fixed income breakdown, stock intersection, top holdings, or returns data.
morningstar-screener-tool
ChatGPTFind investments matching specific criteria across: Stocks, Exchange-Traded Funds, Open End Funds, Closed End Funds, Collective Investment Trusts. This tool filters investments based on quantitative attributes using simple comparison operations. All criteria provided are combined with AND logic (all must match). For complex queries requiring OR logic (e.g., rating=4 OR rating=5), make separate tool calls and combine results. Workflow: 1. Use morningstar-id-lookup-tool to get datapoint IDs 2. Call this tool with universe and list of criteria 3. For OR conditions, make multiple calls and merge results Important: Handling Datapoint Values: The morningstar-id-lookup-tool provides a list of possible values for datapoints. - If the list is complete (has_more_values=False): Use an exact match from the list. - If the list is incomplete (has_more_values=True): Use the provided values as a reference and send your best effort guess for the value. Results are automatically paginated with 200 results per page to prevent context overload. Default Sorting: - Stocks (ST), ETFs (FE), Closed-End Funds (FC), Collective Investment Trusts (CZ): Descending by Morningstar Rating - Open-End Funds (FO): Descending by Fund Size Pagination: - Results are limited to 200 per page - If total_result_count > 200, use pagination_token_next with pagination_action="next" to fetch the next page - Use pagination_token_previous with pagination_action="prev" to fetch the previous page - All other parameters (universe, criteria) must remain identical when paginating - The API returns sorted results, so pagination maintains sort order
morningstar-analyst-research-tool
Claudemorningstar-articles-tool
Claudemorningstar-data-tool
Claudemorningstar-fund-holdings-tool
Claudemorningstar-id-lookup-tool
Claudemorningstar-screener-tool
Claude
