query - SKILL.md Agent Skill

name: query description: Compile a semantic query to SQL using airlayer. Use when the user wants to generate SQL from .view.yml schemas, test a query against their semantic layer, or translate dimensions/measures/filters into dialect-specific SQL. allowed-tools: - Bash - Read - Glob - Grep argument-hint: "[--dimension view.dim --measure view.measure --filter view.dim:operator:value]"

airlayer Query — Compile Semantic Queries to SQL

You are compiling a semantic layer query to SQL using the airlayer CLI.

Prerequisites

Ensure airlayer is installed:

which airlayer || cargo install --git https://github.com/oxy-hq/airlayer

Locate the semantic layer

Find .view.yml files in the project. They are typically under a views/ or semantics/views/ directory.
Identify the base directory (parent of views/).

find . -name "*.view.yml" -not -path "*/node_modules/*" -not -path "*/cube/*" 2>/dev/null | head -20

Inspect available members

Before building a query, inspect what's available:

airlayer inspect```

This lists all views, dimensions, measures, and entities.

## Build and run the query

Use CLI flags (preferred for LLM tool use):

```bash
airlayer query -d <dialect> \
  --dimension <view.dimension> \
  --measure <view.measure> \
  --filter <view.dimension>:<operator>:<value> \
  --order <view.member>:desc \
  --limit 100

Or JSON input for complex queries:

airlayer query -d <dialect> -q '{
  "dimensions": ["view.dimension"],
  "measures": ["view.measure"],
  "filters": [{"member": "view.dim", "operator": "equals", "values": ["val"]}],
  "order": [{"id": "view.measure", "desc": true}],
  "limit": 100
}'

Arguments from user

$ARGUMENTS

Dialect selection

Pick the dialect based on the project's database:

postgres (default), mysql, bigquery, snowflake, duckdb, clickhouse, databricks, redshift, sqlite, domo

If the project has a config.yml with database definitions, use -c config.yml instead of -d.

Filter operators

equals, notEquals, contains, notContains, startsWith, notStartsWith, endsWith, notEndsWith, gt, gte, lt, lte, set, notSet, inDateRange, notInDateRange, beforeDate, beforeOrOnDate, afterDate, afterOrOnDate, onTheDate

Advanced features

Segments: --segments view.segment_name (predefined filter conditions)
Join hints: --through entity_name (disambiguate multi-path joins)
Time dimensions (JSON only): "time_dimensions": [{"dimension": "view.date_col", "granularity": "month", "date_range": ["2024-01-01", "2024-12-31"]}]
Relative dates (JSON only): "date_range": ["last 7 days"] or ["this month"]
Ungrouped mode (JSON only): "ungrouped": true for raw rows without aggregation

Output

airlayer prints the generated SQL to stdout and params to stderr. Show the SQL to the user.