ktx Data Agent Context Layer

Skill by ara.so — MCP Skills collection.

ktx is a self-improving context layer that teaches AI agents how to query your data warehouse accurately. It automatically ingests metadata from databases, dbt, LookML, Looker, Metabase, and Notion, builds a semantic layer with approved metric definitions, and exposes everything via CLI and MCP for agent execution.

What ktx Does

• Learns company knowledge: Ingests wiki content, organizes it, removes duplicates, flags contradictions
• Maps the data stack: Samples tables, captures metadata, detects joinable columns, annotates sources
• Builds semantic layer: Combines raw tables and metrics through a join graph that resolves fan/chasm traps
• Serves agents: CLI and MCP tools with semantic search across wiki and semantic-layer entities

Key benefits for agents:

• Query with approved metric definitions instead of inventing SQL every time
• Reuse canonical business logic across questions
• Get context from scattered sources (dbt, Looker, Notion) in one searchable surface

Installation