Kaelio/ktx

Google Drive Doc Synthesis reads staged Google Drive documents and writes durable wiki pages that capture reusable organizational knowledge. Use it when a WorkUnit contains Google Doc raw files from docs/** and needs search, evidence reconciliation, and semantic-layer verification.

Historic SQL Patterns identifies recurring cross-table analytical intents from historic SQL usage shards and emits typed pattern evidence for deterministic wiki projection. Use it when processing a patterns-input shard from the historic-sql adapter.

Historic SQL Table Digest converts a changed historic-SQL table usage bucket into typed table usage evidence for deterministic schema projection. Use it when a WorkUnit raw file is a tables/<schema>.<name>.json from the historic-sql adapter.

Ingest Triage classifies and resolves conflicts detected during data bundle ingest, including structural duplicates, definitional contradictions, near-duplicate clusters, re-ingest changes, and evictions. Use it before writes that could collide with existing artifacts in a semantic layer.

KTX Analytics queries a KTX-connected database using MCP tools for discovery, semantic-layer analysis, and raw read-only SQL to investigate metrics, explore tables, compare periods, and explain findings. Use it for any data-analysis request against a configured KTX connection.

Ktx installs and configures the open-source context layer for data agents. Use when an agent needs to add ktx to a project, connect data sources, install agent rules, ingest schema, or troubleshoot a local setup.

Live Database Ingest creates or updates semantic-layer sources from a live database schema snapshot, preserving table and column comments. Use it when an ingest work unit contains raw database table JSON under raw-sources.

LookML Ingest maps LookML views, models, and explores into ktx semantic layer sources, covering overlays, standalones, provenance tagging, and decision rules. Use it when a task contains .lkml content or requires writing semantic layer sources from LookML.

Looker Runtime Ingest extracts durable knowledge from Looker dashboards, Looks, and explores. Use it when WorkUnits contain raw files under explores/, dashboards/, or looks/ to derive semantic-layer contributions and wiki knowledge.

Metabase Ingest converts Metabase questions, models, and metrics into ktx Semantic Layer source definitions. It maps result metadata to KSL column types, detects PK/FK relationships, deduplicates near-duplicates, decomposes pre-aggregations, and handles join-graph connectivity.

MetricFlow Ingest maps MetricFlow semantic_models and metrics into ktx semantic layer sources, flattening inheritance and translating metric types like derived, ratio, cumulative, and conversion. Use it when working with .yml/.yaml files containing top-level semantic_models or metrics.

Notion Synthesis reads staged Notion pages, databases, and evidence chunks to write durable wiki pages and semantic-layer sources for business knowledge. Use it when a WorkUnit contains Notion raw files or clustered Notion evidence.

Semantic Layer provides a structured YAML catalog of sources, measures, joins, and segments for querying predefined business metrics like ARR, churn, retention, LTV, and MAU. Use it to read SL source YAML or query metrics via sl_query.

Semantic Layer Capture defines when and how to capture new reusable patterns into ktx's semantic layer—measures, segments, computed dimensions, and joins—so future queries can reuse them instead of rediscovering logic. Use it after a turn produces a reusable business metric or analytical view.

Sigma Ingest extracts durable ktx wiki knowledge from staged Sigma data model specs and workbook summaries. Run it for WorkUnits with unitKey sigma-data-models or sigma-workbooks.

Wiki Capture records durable business knowledge like metric definitions and organizational conventions so an analytics assistant can reuse them across chats. Use it to capture user preferences, rules, and mappings that are not already in the semantic layer.

dbt Ingest maps dbt schema.yml and properties.yml models and sources into ktx semantic-layer overlays and column notes. It handles sources vs models, column data_tests, and resolves physical identifiers before writing semantic-layer sources.