The Databook MCP server lets your AI applications and agents use Databook capabilities directly, without you having to build against the API. It implements the Model Context Protocol (MCP), an emerging open standard for connecting AI clients to external tools, so any client that supports MCP can connect to it.
The MCP server exposes a defined set of tools, not general access to Databook. Only the capabilities described on this page are available through it.
What it exposes
The tools fall into three groups.
Conversational loop Lets you talk to DatabookAI, and to Databook coaches and agents, from inside your AI client. This is the broadest of the three: you can ask any question you could ask DatabookAI, and run coaches to produce assets.
Pattern insight tools Purpose-built tools for the questions Databook users ask most often. Because each one follows a fixed pattern, it answers faster and more reliably than routing the same question through the general conversational loop.
Tool | What it returns |
Company Summary | A concise overview of a company's business model |
Company Overview (Business) | Recent financial performance by business segment and geographic region |
Business and Financial Challenges | A summary of the challenges facing a given company |
Negative Financials News | News from the last six months on negative financial trends |
Earnings Call Summary | A summary of a company's latest earnings call |
Earnings Transcript Issues | The key issues analysts raised in the last two earnings transcripts |
Operational Metrics | Whether, and which, operational metrics a company cites in its earnings transcripts |
Key Developments Search | Key developments in the last twelve months, including executive changes, news, partnerships, and M&A |
Management Initiatives Query | The key management initiatives underway at a company |
Management Intent (Digital) | The latest management intent on digital transformation |
Investment Query | Whether a company is investing in a specific technology |
Industry Financial Performance | Industry-wide financial performance across a company's industry |
Industry Strategic Priorities | The strategic priorities shared across a company's sector |
Strategic Priorities Summary | How a named person helps a company achieve its strategic priorities |
Product Alignment (Strategic) | How your product capabilities align with an account's strategic priorities |
Product Alignment (Financial) | How your capabilities align with an account's business and financial challenges |
Product Alignment (Operational) | How your capabilities align with an account's operational initiatives |
Product Alignment (Digital) | How your capabilities align with an account's digital priorities |
Product Alignment (Sustainability) | How your capabilities align with an account's sustainability priorities |
Job result retrieval Several tools run asynchronously, so a separate tool retrieves the result once a job has finished. Your AI client handles this automatically. You'll typically just notice that some answers take longer than others.
Example use cases
Account research inside your AI assistant: ask about an account and get Databook intelligence back, in the same conversation as everything else you're working on
Running a coach outside Databook: run a Databook coach in Claude and get the finished brief or deck without switching applications
Meeting preparation: pull earnings summaries, analyst issues, and recent developments together ahead of a call
Value story building: use the product alignment tools to connect your own capabilities to a target account's priorities
Agentic workflows: have an agent call Databook as one step in a longer piece of work it's doing on your behalf
What the MCP server is not
It's not general access to Databook's capabilities or data. Only the tools listed above are available.
It doesn't expose Databook's wider internal toolset.
It's not a route to bulk extraction of raw data.
It's not a substitute for the Batch or Reasoning API. If you need high-volume, scheduled generation across many accounts, use that instead.
How it will change
MCP is an emerging standard, and the Databook server will change as it matures. Tools, resources, and prompts may be added, changed, or withdrawn. The same capabilities are available to every customer; what differs is how your own Databook instance is configured.
Conditions of use
The same restrictions that apply to the API apply to the MCP server:
No using output to train, fine-tune, benchmark, or validate any AI or machine learning model, including by the provider of the AI application you're connecting
Disclaimer and citation requirements must be displayed wherever output is surfaced, including inside an AI client, a chat transcript, a document, or a deck
No extracting, caching, aggregating, or reconstructing raw underlying data
You're responsible for satisfying yourself about how your AI application provider and its model provider process output and your data
Databook can throttle or suspend access that exceeds documented limits or threatens the stability of the service.
Getting connected
See Connecting to the Databook MCP server for configuration values and step-by-step setup.