metabase-semantic-checker
Runs the Metabase semantic checker against a tree of Representation Format YAML files to verify that all references resolve — cross-entity references (collectionid, dashboardid, parentid, parameter source cards, snippet references, transform tags, etc.) and references to columns inside MBQL and native queries. Slow (≥1 min per run). Only use when the user explicitly asks to verify entity references or column references in MBQL/SQL queries; in most cases this runs as a CI step, not locally. Requires database metadata on disk (by default .metadata/tablemetadata.json).
- Installs
- 434
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/metabase/agent-skills --skill metabase-semantic-checker- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads metabase-semantic-checker automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Metabase semantic checker
The semantic checker validates a tree of Metabase Representation Format YAML files for referential integrity. Schema-level validation (shape of each file, required fields, enum values) is handled separately by npx @metabase/representations validate-schema; the semantic checker runs after schema validation and focuses on cross-file and cross-system consistency.
It compiles every MBQL query down to SQL against the database metadata and checks that each entity reference and each column reference resolves to something that actually exists. Concretely, it answers:
- Does every
collection_id,parent_id,dashboard_id,document_id,based_on_card_id, transform tag, snippet name, etc. resolve to an entity that actually exists in the tree? - For each MBQL query, do every
source-table, field reference, join target, segment, measure, and expression resolve against the database schema? (Verified by compiling the query to SQL.) - For each native query, do the referenced tables, columns, and snippets exist?
- Do dashboards' and documents' embedded card references point at real cards?
Each run takes 1 minute or more — roughly a minute of fixed JVM + metadata-loading overhead before any checks start, plus query-compilation time that scales with the tree.
The checker ships inside the Metabase Enterprise JAR and is invoked via --mode checker. Default Docker image: metabase/metabase-enterprise:latest. Use metabase/metabase-enterprise-head:latest only when the user explicitly wants the in-development build — e.g. testing unreleased checker changes.
Inputs
Two inputs, both required:
- The representation tree — the repo root containing
collections/,databases/,transforms/,python_libraries/. This is what gets checked. - The database metadata — a JSON file exported from a Metabase instance. By default located at
.metadata/table_metadata.json. The checker uses it to resolve column/table references inside queries; without it, query-level checks cannot run.
If .metadata/table_metadata.json is missing, do not run the checker. Tell the user it needs to be exported from their Metabase instance first, and only run the checker once the metadata file is present on disk.
When to run
Do not run the semantic checker by default when making edits. It is slow (≥1 minute per run) and in most projects is wired up as a CI step that runs on every push or PR — that is where it belongs. Local runs are for targeted diagnosis, not routine validation.
Only run it locally when the user explicitly asks for one of these:
- verify that all entity references resolve (collections, dashboards, cards, snippets, transform tags, etc.), or
- verify that all column references in queries — MBQL or SQL — are correct.
Phrasings that count as an explicit ask: "semantic check", "check references", "validate queries against the schema", "make sure the columns still exist", or diagnosing a broken reference the user already suspects. A bare "run the checker" does not count — by default "the checker" means the fast schema checker (npx @metabase/representations validate-schema). Only wording that explicitly names references or queries should trigger the semantic checker.
Otherwise, skip it. After editing YAML, rely on npx @metabase/representations validate-schema for local feedback and leave the semantic check to CI. Do not run it proactively at session start, and do not run it as a self-imposed "finishing step" after edits unless the user asked for it.
If you do run it, batch. Make all the YAML changes first, then run the checker once. Each invocation pays the ≥1-minute fixed overhead; running between edits multiplies that cost. If it surfaces issues, fix everything you can see in one pass before re-running.
Running the checker
Once .metadata/table_metadata.json exists and Docker is available:
docker pull metabase/metabase-enterprise:latest
docker run --rm \
-v "$PWD:/workspace" \
--entrypoint "" \
-w /app \
metabase/metabase-enterprise:latest \
java -jar metabase.jar \
--mode checker \
--export /workspace \
--schema-dir /workspace/.metadata/table_metadata.json \
--schema-format concise
Flag reference:
--mode checker— selects semantic-check mode (skips server startup, import, etc.).--export /workspace— path inside the container to the representation tree root. With the-v "$PWD:/workspace"mount above, this maps to the current repo root on the host.--schema-dir /workspace/.metadata/table_metadata.json— path to the database metadata JSON. Despite the-dirsuffix the flag accepts a single JSON file. Point it elsewhere only if the user has stored metadata at a non-default path.--schema-format concise— format the input metadata is in.concisematches what a Metabase instance exports. Do not change unless the user explicitly has a different dump format.
The container needs no network access for the check itself — pull the image first if the host is offline-prone.
Exit code is non-zero on findings. Surface the checker's stdout/stderr verbatim to the user; do not summarize away specific paths or entity names, since those are how the user locates the broken reference.
Common failure modes
- "Database metadata not found" / schema load errors —
.metadata/table_metadata.jsonis missing, stale, or malformed. Ask the user to re-export it from their Metabase instance. - Unknown collection / card / dashboard / snippet / tag reference — the referenced
entity_idor name does not exist in the tree. Either the target YAML is missing, or the reference is a typo; grep the tree for the id/name to confirm which. - Unknown table or field inside a query — the query references a column that the database metadata doesn't know about. Either the warehouse schema has drifted (refetch metadata), or the query itself is wrong.
- Docker image missing / not pulled — run
docker pull metabase/metabase-enterprise:latestfirst. On slow networks warn the user; the image is multi-hundred-MB.
FAQ
What does the metabase-semantic-checker skill do?
Runs the Metabase semantic checker against a tree of Representation Format YAML files to verify that all references resolve — cross-entity references (collectionid, dashboardid, parentid, parameter source cards, snippet references, transform tags, etc.) and references to columns inside MBQL and native queries. Slow (≥1 min per run). Only use when the user explicitly asks to verify entity references or column references in MBQL/SQL queries; in most cases this runs as a CI step, not locally. Requires database metadata on disk (by default .metadata/tablemetadata.json).
How do I install metabase-semantic-checker?
Run npx skills add https://github.com/metabase/agent-skills --skill metabase-semantic-checker in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/metabase-semantic-checker/ (personal) or .claude/skills/metabase-semantic-checker/ (project).
Which AI agents can use metabase-semantic-checker?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is metabase-semantic-checker safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains metabase-semantic-checker?
It is published by metabase in metabase/agent-skills. It has 434 recorded installs.
More from metabase/agent-skills
rde
metabase/agent-skills
Expert data engineering with Metabase at the center of any company's stack: explore and profile raw warehouse tables, build clean tables as Metabase transforms pinned by transform tests, build the semantic layer (models, metrics, measures, segments, metadata), design dashboards, answer questions with checked numbers, reconcile against a reference, change delivered definitions safely. Use for data work, not one mb command (metabase-cli): "make sense of my data", "model this raw schema", "set up analytics for X", "build a data model", "define MRR / active customers officially", "build a semantic layer", "go from raw tables to a dashboard", "does this number match finance", "our numbers look wrong", "two cards disagree", "test this transform", "why does this transform get case X wrong", "make Metabot answer questions about X", "explain this table", "change this definition", "build a data app", "donor retention", "event attendance", "quarterly board report", "be my data engineer / anal
rde-data-apps
metabase/agent-skills
Build or change a Metabase data app on an rde machine: settles what the metabase-data-apps skill needs from the local setup (the remote-sync repository, the Metabase URL, the API key rde stores, how the app reaches Metabase), makes sure everything the app reads is Library content already synced into that repository, confirms every sync step with the user, and hands the build to it. Use only when an app is asked for explicitly: "build me an app for X", "create a data app", "make an internal tool / portal in Metabase", "add a page / a form / a filter to my data app". A dashboard, a question, or a metric is not an app; those stay with the rde skill.
metabase-cli
metabase/agent-skills
Drive a Metabase instance from the terminal via the mb CLI: auth profiles, databases and schemas, schema sync and field-values rescan, tables, fields, cards (questions, models, metrics) run as JSON/CSV/XLSX, dashboards and dashcards, collections and their tree, snippets, segments, measures, transforms and transform-jobs, settings, search, git-sync. Use for any Metabase operation from the terminal: "log into metabase", "what profiles do I have", "list cards", "run card 42 as CSV", "create a transform", "list dashboards", "move a dashcard", "list collections", "what's in collection 4", "show the collection tree", "list snippets", "create a segment", "archive a measure", "search metabase for X", "import the latest changes", "add a directory to git sync", "set a setting", "what schemas are in this database", "trigger a sync", "rescan field values", or any mb . Whole data jobs go to rde.
metabase-database-metadata
metabase/agent-skills
Understands the Metabase Database Metadata Format — a YAML-based on-disk representation of databases, tables, and fields synced from a Metabase instance. Use when the user needs to read, edit, or understand metadata files produced by @metabase/database-metadata, or when reasoning about a project's schema (columns, types, FK relationships) through the .metadata/databases folder.
metabase-representation-format
metabase/agent-skills
Understands the Metabase Representation Format — a YAML-based serialization format for Metabase content (collections, cards, dashboards, documents, segments, measures, snippets, transforms, actions, and data apps). Use when the user needs to create, edit, understand, or validate Metabase representation YAML files, or when working with Metabase serialization/deserialization (serdes). Covers entity schemas, MBQL and native queries, visualization settings, parameters, and folder structure.
metabase-modular-embedding-version-upgrade
metabase/agent-skills
Upgrades a project's Metabase Modular embedding SDK (@metabase/embedding-sdk-react) or Modular embedding (embed.js) version. Use when the user wants to upgrade their Metabase modular embedding integration to a newer version.
Popular data & databases skills
supabase-postgres-best-practices
supabase/agent-skills
Postgres best practices maintained by Supabase, for Postgres running anywhere. Load this skill BEFORE writing or changing anything that lives in a Postgres database: creating or altering tables and columns (including choosing column types), schema design, migrations and declarative schema files, RLS policies and the tests that verify them, indexes, triggers, database functions, queues and scheduled jobs (pgcron, pgmq), vector/semantic search (pgvector), and restoring dumps (pgrestore) or importing data. Also load it when diagnosing slow queries, high CPU, timeouts, EXPLAIN plans, connection exhaustion, locking, bloat, or rows visible to the wrong user or tenant. This is not just a performance guide — schema, migration, security, and SQL authoring tasks need these rules too, even for a one-column change or a single query.
prisma-client-api
prisma/skills
Prisma Client API reference covering model queries, filters, operators, and client methods. Use when writing database queries, using CRUD operations, filtering data, or configuring Prisma Client. Triggers on "prisma query", "findMany", "create", "update", "delete", "$transaction".
prisma-database-setup
prisma/skills
Deprecated compatibility name for prisma-orm-setup. Use only when a user or existing integration explicitly invokes prisma-database-setup; otherwise select prisma-orm-setup.
prisma-cli
prisma/skills
Prisma ORM 7 CLI reference for existing Prisma 7 applications or explicitly selected Prisma 7 workflows. Covers prisma init, generate, migrate, db, dev, studio, and mcp. For setup, connection repair, or Prisma 8, use prisma-orm-setup; not the Prisma Platform CLI.
prisma-postgres
prisma/skills
Deprecated compatibility name for prisma-postgres-setup. Use only when a user or existing integration explicitly invokes prisma-postgres; otherwise select prisma-postgres-setup.
prisma-driver-adapter-implementation
prisma/skills
Required reference for Prisma ORM 7 SQL driver adapter work. Use when implementing or modifying adapters, adding database drivers, or touching SqlDriverAdapter, Transaction, savepoint, result mapping, or DriverAdapterError behavior. Covers current transaction lifecycle, optional savepoint hooks, original database-error preservation, and verification.