# Morningstar Agent
Discover emerging trends, optimize investment strategies, and identify risks and opportunities with AI. {.sub-title___devsite}
## Overview
The Morningstar Agent brings Morningstar's proprietary financial data, research, and analytics directly into AI-powered workflows and conversations. It is a managed financial AI system built specifically for investment and financial analysis workflows, combining Morningstar's proprietary content with purpose-built orchestration, conversational memory, contextual grounding, and enterprise governance controls.
Using the Morningstar Agent, you can:
- Look up stocks, ETFs, and mutual funds by name, ticker, or ISIN
- Access hundreds of structured datapoints, including valuations, performance metrics, risk statistics, ratings, and historical data
- Retrieve Morningstar analyst research covering economic moat analysis, fair value estimates, business strategy assessments, and Bulls Say/Bears Say perspectives
- Screen universes of stocks, ETFs, and funds against custom investment criteria
- Analyze fund holdings and portfolio composition
- Search Morningstar's editorial article library for market commentary and investing insights
Unlike general-purpose AI assistants, the Morningstar Agent is optimized for financial workflows. It dynamically orchestrates Morningstar MCP tools, maintains conversational context across interactions, grounds responses in Morningstar content, and returns traceable outputs with clear data lineage.
> Access to the Morningstar Agent requires an active Morningstar license.
## Developer Integration
Developers access the agent through a unified API endpoint. There is no need to build custom retrieval pipelines, orchestrate tool execution loops, engineer financial-domain system prompts, or maintain context memory infrastructure.
Once you have a Morningstar account, you can generate an authentication token using the
[Authentication endpoint](/direct-web-services/documentation/api-utilities/authentication-api/overview).
### Example cURL
```bash
curl -X 'POST' \
'https://agents.morningstar.com/morningstar-agent/v1/run' \
-H 'accept: application/json' \
-H 'Authorization: Bearer ' \
-H 'Content-Type: application/json' \
-d '{
"message": "What is the current outlook of the US economy?"
}'
```
## Why the Morningstar Agent?
General-purpose AI systems are optimized for broad knowledge tasks. Investment workflows require stronger guarantees around data quality, grounding, consistency, context retention, financial-domain reasoning, and governance. The Morningstar Agent is built to meet those requirements out of the box.
### How the Morningstar Agent Works
The Morningstar Agent combines:
- A large language model
- Morningstar MCP
- Domain-specific orchestration through specific system prompts
- Context management infrastructure
- Conversational memory
- Safety and grounding systems
Unlike raw MCP integrations, developers do not need to manually orchestrate tool execution loops or manage context windows.
### How Does Conversation Memory Work?
Every response includes an `X-Context-ID` header. Passing this header in subsequent requests tells the agent to maintain conversational context, keeping a history of questions and responses available across conversational turns.
### Morningstar Agent vs. Morningstar MCP Server
| | **Morningstar MCP Server** | **Morningstar Agent** |
|---|---|---|
| **Access** | Direct tool and data access through MCP Clients like Claude, ChatGPT, Perplexity | Managed Morningstar API endpoint |
| **Orchestration** | Developer or client-owned | Fully managed |
| **Context and memory** | Developer or client-owned | Built-in |
| **Governance and safety** | Developer or client-owned | Built-in |
| **Auditability** | Developer or client-owned | Available on demand |
| **Best for** | Teams building custom AI orchestration | Teams integrating a production-ready financial AI experience |
To learn more about our MCP server, see the
[Morningstar MCP Server](/direct-web-services/documentation/morningstar-ai-integrations/morningstar-mcp-server/overview) documentation.
### Where Can I Use This Agent?
The Morningstar Agent can be embedded into existing workflows or used as a conversational chatbot that maintains context across multiple turns.
**As a chatbot**, it supports natural, multi-turn financial conversations. For example:
- *"What is the fair value of Apple?"* followed by *"How does that compare to its current price?"* — the agent understands that "that" refers to Apple's fair value from the previous turn.
- *"Which large-cap tech stocks have a wide moat?"* followed by *"Which of those are currently undervalued?"* — no need to restate the original criteria.
**As part of a workflow**, it can be integrated into research pipelines, client-facing tools, or internal analyst platforms where financial data and AI-generated insights need to work together seamlessly.
## Safety and Governance
Access and entitlements are strictly enforced using JWT tokens, with individual user accounts determining access to specific Morningstar data and research.
User access is controlled through existing data packages powered by our backend APIs, with entitlements managed automatically so you don't need to handle access orchestration yourself.
The agent also includes:
- PII detection and redaction
- Prompt injection mitigation
- Grounded response enforcement
- Auditability
## Related Content
- [API Reference](/direct-web-services/documentation/morningstar-ai-integrations/morningstar-agent/api-reference)