whereismyai
Ontology Skills
Ontology Skills: Skills to bootstrap and expand analytics ontologies from schema, dbt, dashboards and docs.
Ontology Skills builds and expands the context an analytics agent needs by starting from existing data assets and attaching evidence to definitions. The kit uses a schema export, a repository that builds tables such as a dbt project or SQLMesh, optional column glossaries, warehouse comments, documentation, dashboard exports carrying SQL, semantic layers, and query logs. A run produces a tree of reviewable files, metrics with corroboration stamps, and blocking questions that require human input. It leaves blanks rather than guessing when nothing supports a definition, and a verification pass rereads each claim against its source. It is designed for data teams who need to bootstrap or grow analytics ontologies while keeping control over business meaning.
Ontology Skills at a glance
Key features
- Builds initial ontologies from existing data stacks
- Expands previously maintained ontologies
- Produces reviewable file trees
- Leaves blanks and files questions when evidence is missing
- Verifies claims against named sources
- Provides four review checkpoints for human decisions
- Integrates schema exports and dbt projects directly
Use it for
- Bootstrapping a first ontology from existing data infrastructure
- Expanding an already maintained ontology
- Corroborating metrics using dashboards and query logs
- Identifying pipeline defects and contradictory documentation
Alternatives to Ontology Skills
- ktx — It gives data agents a shared context workspace of approved metrics and git-based files.
- Evidence — It defines business intelligence as code, enabling collaboration and automation.
- Portia — Computes every number with code so claims are verifiable.
- Bricks — It creates professional data dashboards from spreadsheets using natural language.
- yggy — Keeps knowledge graphs automatically up to date with source citations for every fact.
- DB-GPT — Builds autonomous agents for complex data tasks with private LLM technology.
- Metabase — Query and visualize data from any database using plain English, SQL, or clicks.
- Upsolve — Grounds AI agent answers in versioned data models and definitions.
- Flint — Compiles compact specifications into native output for multiple chart libraries.
- Cube — It keeps metric definitions consistent across AI agents and human users.
- Helix Foundry — Connects company data into a local ontology with local-first AI and no cloud accounts.
- Jonex — Provides traceable, agent-ready knowledge from diverse enterprise data sources.
All Ontology Skills alternatives