Scout MCP
Good dashboards require a strong understanding of the data. Finding the right sources is the first hurdle of any analysis.
Scout MCP connects your team’s compatible AI tools to the sources and context they need to explore. It’s designed to be token-light, bringing in relevant context rather than loading everything into each conversation. Find the relevant tables, understand the fields and check the logic before building the report.
Included in your Plotline implementationGet to know your data.
Start with a question and work back to the source. Scout helps your team make sense of what’s available.
Find the right sources
Explore connected tables, views and documentation to find a useful starting point.
Understand the context
Inspect schemas, definitions and documented relationships before joining things together.
Check the logic
Develop queries with the source and calculation visible, so your team can review the result.
Use shorter prompts
When you use agents to write queries, Scout MCP helps them discover the relevant data sources. Describe what you want to analyse, without having to list the tables and fields in every prompt.
Where does revenue live?
Follow the question back to its source.
YouWhich tables should I use for monthly revenue?
finance.monthly_revenueMonthly recognised revenue, by account.Start here for the finance report.sales.ordersIndividual orders and their status.Use for order activity, rather than recognised revenue.YouWhat does one row represent?
finance.monthly_revenueOne account, one month, one reporting currency.- month
- Revenue recognition month
- account_id
- Account identifier
- recognised_gbp
- Recognised revenue in GBP
These definitions would come from the documentation made available to Scout.
YouShow me the logic for monthly revenue.
SELECT month,
SUM(recognised_gbp) AS revenue_gbp
FROM finance.monthly_revenue
GROUP BY month
ORDER BY month;Check the definition with finance before using it in a dashboard. Order value and recognised revenue answer different questions.
Illustrative workflow. This example is not connected to a database.
Part of your implementation.
We use Scout to help understand the data behind your reporting. The agreed setup comes with your Plotline implementation, so your team can keep exploring after handover.
You receive the code, configuration and documentation. There’s no separate required Scout subscription.
What we agree with you
- Sources
- The databases and documentation to connect.
- Access
- The users, permissions and query scope.
- AI tools
- The compatible client and provider your team chooses.
- Handover
- Code, configuration and documentation.
AI-provider usage, hosting and optional maintenance are scoped separately.
Practical questions
What is MCP?
Model Context Protocol is a way for compatible AI applications to work with connected tools and data sources. Scout is the MCP server we configure around the sources and context agreed for your implementation.
Can it access everything in our database?
Access should be limited to what your team approves. We agree the connections and scope during setup and configure the underlying access accordingly. MCP itself does not grant database permissions.
Where does our data go?
That depends on where Scout runs and which AI client and provider you connect. We agree what may be accessed and sent to the provider before connecting company data.
Is Scout an extra subscription?
The agreed Scout setup is included in your Plotline implementation price. Your company pays any hosting and AI-provider costs. Optional maintenance and further connections are scoped separately.
Want to use Scout with your data?
Tell us about your sources and what your team needs to understand.