Available tools in the Yogi MCP | Yogi Knowledge Base

The Yogi MCP server exposes a set of tools your AI assistant can call on your behalf to answer questions about your consumer feedback data. You don't need to know which tool to use — your assistant picks the right one, or combines several, based on what you ask. This page lists what each tool does, so you can see what Yogi can answer and what it can't.

All tools respect your Yogi permissions. You won't see any data through your assistant that you couldn't see in the Yogi app.

Understanding your data

These tools are used in the background to load your hub's context before it answers a question. You won't usually call them directly.

Tool	What it does
`get_analysis_context`	Loads Yogi's analysis methodology for the connected hub — the base analysis system prompt plus any hub-specific guidelines your team has set. Claude calls this once at the start of a conversation so its answers stay consistent with how the Yogi app analyzes your data.
`lookup_filter_options`	Discovers which filter fields exist for your hub (products, sources, sentiment, themes, brands, retailers, custom filters, timeframes, etc.) and fuzzy-matches values you mention in a prompt. If you say "double stuffed Oreos," this is the tool that resolves your natural language into the canonical product name in your hub.
`get_verbatims`	Pulls individual customer feedback records from Yogi. This is used to surface specific verbatims back to you, and to summarize what customers are saying (since the LLM has to read the records to summarize them).
`aggregate_and_group_by`	Any quantitative answer: counts, averages, sums, min/max, grouped by any dimension (product, theme, retailer, sentiment, date bucket, etc.). This is the tool behind trends, breakdowns, and comparisons.
`create_chart`	Turns your data into presentation-ready charts — bar, line, pie, stacked bar, scatter, or bubble — for comparisons, rankings, trends, distributions, and correlations. This is used when a visual will land the answer better than a table. Every chart comes back with a link that opens the same view in the Yogi.

Coming soon

Additional analysis tools — surfacing signals from the Issue Detection Agent and Benchmarking Agent as callable tools.
Write actions — creating and sharing dashboards, drafting alerts, and spinning up agents from Claude.

Example: How Claude uses these tools

You don't call these tools directly — Claude does, based on your prompt. A typical flow:

You ask a question in plain language.
Claude loads get_analysis_context to understand your hub's methodology.
Claude calls lookup_filter_options to resolve any product names, sources, or filters you mentioned.
Claude calls get_verbatims, aggregate_and_group_by, or both (sometimes several times with different filters) to answer.
When a visual will make the answer clearer, Claude also calls create_chart so the chart shows up alongside the written analysis, with a link to open it in Yogi.
Claude returns the answer and links back to the app for deeper work.