Build an AI Stock Analyst with Fred MCP

Learn how to build an AI stock analysis like Robinhood Cortex using Fred MCP.

Woojong Koh

October 23, 2025 · 5 min read

Fred MCP is a service designed to transform any Large Language Model (LLM) into an expert AI stock analyst.

What is an MCP? #

You might be wondering, what is an MCP?

An MCP (Model Context Protocol) server is an application that exposes tools and resources to an LLM through the Model Context Protocol, allowing it to interact with external systems and data sources.

Think of MCP as a next-generation API standard for the new AI era, specifically for building AI agents. The first and most crucial step toward a powerful AI agent is an augmented LLM.

The difference is night and day.

An LLM is like a brilliant student taking an exam in an isolated room. They only know what they were trained on.
An LLM with Fred MCP is like that same student with a smartphone, granting them real-time access to professional-grade, LLM-friendly financial information.

For analyzing U.S. stock price movements and predicting their prospects, this connection to real-time, comprehensive data is a game-changer.

Case Study: Robinhood Cortex Digests #

Robinhood Cortex Digests are insights that leverage artificial intelligence (AI) to generate plain-language summaries of what may be driving a stock’s price.

This is exactly the type of service you can build. With Fred MCP, you can replicate the core functionality of a service like Cortex Digests in minutes.

Data We Provide #

Currently, Fred MCP offers the following datasets, all formatted for an LLM:

SEC Filings: Comprehensive access to 10-K, 10-Q, and 8-K reports.
Corporate Earnings Calls: Full audio recordings and presentation slides.
Financial News: The latest stock market news from major U.S. media outlets.
Social Media: User posts from Reddit and Truth Social.
Podcasts: Audio from financial and business podcasts (U.S. and U.K.).
Market Data: Real-time U.S. stock prices and a wide array of technical analysis indicators (SMA, EMA, RSI, MACD, Bollinger Bands, On-Balance Volume).
Proprietary Data: Our own extensive corporate Knowledge Base (KB) and ontology, built by Fred.

Example: Prompt and Response #

Here is an example of a prompt you can send to an LLM that is connected to the Fred MCP.

Prompt:

Analyze the price movement of PLTR stock over recent trading days.
Use all available information, including SEC filings, stock news, earnings call transcripts, Reddit posts, knowledge base documents, stock prices, and technical indicators.
Write the analysis in the following format in both English and Korean:
For English:
Please write in a polite, informal style.
Examples:
Importance: High. A new CEO has been appointed. They’re reportedly the former CFO from during Netflix’s +10,000% stock price surge. Source: …
Importance: Medium. Financial health is worsening, and the deficit is growing. This was due to the rise in international energy prices and difficulties in adjusting electricity sales prices. Source: …
For Korean:
두루낮춤 비격식체인 해요체로 작성해주세요.
예시:
중요도: 높음. 신규 CEO가 취임했어요. 넷플릭스 주가 연상승률 +10,000% 당시 CFO 출신이래요. 출처: …
중요도: 중간. 재무 건전성이 악화되어 적자가 늘어나고 있어요. 국제 에너지 가격 상승과 전력 판매 단가 조정의 어려움 때문이었어요. 출처: …

Response (Gemini CLI with Gemini Pro 2.5):

English:

Importance: High. Palantir recently announced a strategic partnership with Snowflake, a major player in the data cloud market. This collaboration will integrate Palantir’s Foundry and AIP platforms with Snowflake’s AI Data Cloud, which is expected to drive more usage and demand for both companies’ services. This news has been received positively by the market and is a significant factor in the stock’s recent performance. Source: Reddit post “Snowflake and Palantir Announce Strategic AI Partnership”
Importance: High. The company reported record-breaking earnings for Q2 2025, surpassing $1 billion in quarterly revenue for the first time. U.S. commercial revenue grew by 93% year-over-year, and the company raised its full-year revenue guidance. This strong financial performance has been a primary driver of the stock’s upward trend. Source: Palantir Q2 2025 Earnings Call, SEC Form 8-K filed on August 4, 2025.
Importance: Medium. Technical Analysis: The technical indicators are mostly bullish but show some signs of slowing momentum. The Bollinger Bands are wide, indicating volatility, with the price in the upper half suggesting some upward momentum. The MACD line is above the signal line (a bullish sign), but the histogram is decreasing, which could indicate the upward trend is losing strength. The On-Balance Volume (OBV) is in an uptrend, confirming the price trend. The Relative Strength Index (RSI) is neutral at 53.61, meaning the stock is neither overbought nor oversold. Source: Gemini Technical Analysis Tools.
Importance: Medium. Several financial news outlets have raised concerns about Palantir’s high valuation. The stock is trading at a very high price-to-earnings (P/E) ratio, which some analysts believe is unsustainable and makes the stock vulnerable to a correction. This cautious sentiment from analysts may be contributing to some of the recent price volatility. Source: Quartz article “3 stock market stars that could burn out fast”, The Motley Fool article “Palantir Stock vs. Microsoft Stock: Wall Street Says Only 1 Will Head Higher From Here”.
Importance: Medium. There is a Reddit post pointing to other technical indicators that suggest a potential for short-term weakness in the stock price. While not a fundamental factor, technical analysis can influence short-term trading behavior and contribute to price fluctuations. Source: Reddit post “PLTR - Technical bearish in the Short run”.
Importance: Low. Palantir’s CEO, Alex Karp, has a unique and outspoken communication style that often generates media attention. While his “spicy” shareholder letters and comments on competitors create buzz, they are unlikely to have a direct, lasting impact on the stock’s price movement compared to financial results and strategic partnerships. Source: Fortune article “How Palantir’s CEO forged a connection with investors by writing spicy shareholder letters…”.

Korean:

중요도: 높음. Palantir는 최근 데이터 클라우드 시장의 주요 업체인 Snowflake와 전략적 파트너십을 발표했어요. 이번 협력으로 Palantir의 Foundry 및 AIP 플랫폼이 Snowflake의 AI 데이터 클라우드와 통합되어 양사 서비스에 대한 사용량과 수요가 증가할 것으로 예상돼요. 이 소식은 시장에서 긍정적으로 받아들여졌고 최근 주가 움직임에 중요한 요인이 되고 있어요. 출처: 레딧 게시물 “Snowflake and Palantir Announce Strategic AI Partnership”
중요도: 높음. 2025년 2분기에 사상 최대 실적을 발표하며 분기 매출 10억 달러를 처음으로 돌파했어요. 미국 내 상업 부문 매출은 전년 동기 대비 93% 증가했으며, 연간 매출 가이던스도 상향 조정했어요. 이러한 강력한 재무 성과는 주가 상승의 주된 동력이었어요. 출처: Palantir 2025년 2분기 실적 발표, 2025년 8월 4일자 SEC 8-K 보고서.
중요도: 중간. 기술적 지표는 대부분 강세이지만 일부 모멘텀 둔화 신호가 보여요. 볼린저 밴드가 넓어 변동성이 크다는 것을 나타내며, 가격이 밴드 상단에 위치해 상승 모멘텀이 있음을 시사해요. MACD 선이 시그널 선 위에 있어 강세 신호이지만, 히스토그램이 감소하고 있어 상승 추세가 약화될 수 있음을 나타낼 수 있어요. OBV(On-Balance Volume)는 상승 추세로 거래량이 가격 추세를 뒷받침하고 있어요. RSI(상대강도지수)는 53.61로 중립 구간에 있어 과매수 또는 과매도 상태가 아니에요.
중요도: 중간. 여러 금융 뉴스 매체에서 Palantir의 높은 밸류에이션에 대한 우려를 제기했어요. 주가가 매우 높은 주가수익비율(P/E)로 거래��고 있어 일부 분석가들은 이것이 지속 불가능하며 주가 조정에 취약할 수 있다고 보고 있어요. 분석가들의 이러한 신중한 시각이 최근의 주가 변동성에 일부 기여하고 있을 수 있어요. 출처: Quartz 기사 “3 stock market stars that could burn out fast”, The Motley Fool 기사 “Palantir Stock vs. Microsoft Stock: Wall Street Says Only 1 Will Head Higher From Here”.
중요도: 중간. 레딧의 한 게시물은 주가의 단기적인 약세 가능성을 시사하는 다른 기술적 지표를 지적하고 있어요. 근본적인 요인은 아니지만, 기술적 분석은 단기적인 거래 행동에 영향을 미치고 가격 변동에 기여할 수 있어요. 출처: 레딧 게시물 “PLTR - Technical bearish in the Short run”.
중요도: 낮음. Palantir의 CEO인 Alex Karp는 독특하고 직설적인 소통 방식으로 언론의 주목을 자주 받아요. 그의 “매운” 주주 서한과 경쟁사에 대한 발언이 화제가 되기는 하지만, 재무 실적이나 전략적 파트너십에 비해 주가에 직접적이고 지속적인 영향을 미칠 가능성은 낮아요. 출처: Fortune 기사 “How Palantir’s CEO forged a connection with investors by writing spicy shareholder letters…”.

How to Try It Yourself #

Fred MCP supports any LLM (remote, local, managed, or on-prem) that is ready for the Model Context Protocol.

Fred MCP is currently in beta. We are providing test API keys for evaluation. Please contact us to get your test API key.

For Gemini CLI #

We recommend using Gemini CLI with a model like Gemini 2.5 Pro for testing. It excels at this task because it can be configured to allow all tools without repeated permissions and supports a huge context size (1M tokens).

Install Gemini CLI from: https://geminicli.com/
Edit your ~/.gemini/settings.json file.
Add the fred server configuration under mcpServers:

{
  "ui": {
    ...
  },
  "mcpServers": {
    "fred": {
      "httpUrl": "https://fred-test.cwrk.ai/mcp/",
      "trust": true,
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY_HERE"
      }
    }
  },
  "general": {
    ...
  },
  "security": {
    ...
  }
}

(Remember to replace YOUR_API_KEY_HERE with the key we provide.)

For more details, see the Gemini CLI documentation: https://geminicli.com/docs/tools/mcp-server/#configuration-structure

For Claude Code #

Claude models, such as Sonnet 4.5, are also supported. However, be aware of a few limitations: they currently offer a smaller context size (200K tokens) than Gemini 2.5 Pro (1M tokens), and Claude Code may ask for tool-use permission on each run rather than allowing them by default. If you encounter any issues, we recommend trying the Gemini CLI setup.

You can connect Fred as a remote HTTP server.

claude mcp add --transport http fred-test https://fred-test.cwrk.ai/mcp/ \
  --header "Authorization: Bearer YOUR_API_KEY_HERE"

(Remember to replace YOUR_API_KEY_HERE with the key we provide.)

Check out the official Claude documentation for instructions: https://docs.claude.com/en/docs/claude-code/mcp#option-1%3A-add-a-remote-http-server

Join Our Beta Test Today #

Fred MCP is currently in beta. If you want to experience the next frontier of stock analysis powered by LLMs, please contact us anytime through the link below to request access.

Link to Beta Test Application