Disco
Run hypothesis-free pattern discovery on tabular datasets to decide which metrics or segments deserve a real feature bet.
Overview
Disco Discovery Engine is an MCP server for the Validate phase that finds novel, statistically validated patterns in tabular data without fixed hypotheses.
What is this MCP server?
- Disco discovery engine on streamable HTTP at disco.leap-labs.com/mcp
- Hypothesis-free search for novel patterns in tabular data with statistical validation
- Bearer disco_ API key required on Authorization header
- Version 1.0.1; open-source repo leap-laboratories/discovery-engine
- Leap Labs website leap-labs.com for product context
- MCP server version 1.0.1
- Hosted endpoint disco.leap-labs.com/mcp
- API key prefix disco_ in Authorization header
Community signal: 6 GitHub stars.
What problem does it solve?
Founders stare at CSVs and dashboards but miss non-obvious segment and metric relationships because exploration is manual and question-biased.
Who is it for?
Data-curious solo builders with tabular exports who want agent-assisted exploration before committing engineering time.
Skip if: Unstructured text-only products, real-time streaming pipelines, or teams unwilling to use a paid disco_ API key on a hosted service.
What do I get? / Deliverables
Agents can invoke Disco over your tables and return validated pattern candidates you use to tighten scope and prioritization.
- Statistically validated pattern candidates from agent-driven runs
- Clearer validate-scope priorities backed by data structure
- Repeatable remote MCP hook for ongoing analytics exploration
Recommended MCP Servers
Journey fit
Before you over-build dashboards or ML, you need statistically grounded signals about what is actually in your tables. Scope validation benefits from external discovery that narrows what to prototype without a fixed analyst question list.
How it compares
Hosted analytics discovery MCP, not a spreadsheet plugin or generic SQL assistant skill.
Common Questions / FAQ
Who is Disco Discovery Engine for?
Indie SaaS and product builders who have tabular data and want statistically validated pattern leads before building analytics features.
When should I use Disco Discovery Engine?
Use it when scoping a feature, pricing change, or growth experiment and you need hypothesis-free exploration on existing tables.
How do I add Disco Discovery Engine to my agent?
Add remote https://disco.leap-labs.com/mcp with Authorization Bearer plus your disco_ API key in your MCP client settings.