skills.rest
constructorfabric
constructorfabricOFFICIAL
Explore all skills by this author →
WebsiteGitHub

metric-test

Author and validate YAML fixture and pytest specs for data-path metric tests.

10|9|Updated May 22, 2026
Search Tags:#pytest#data pipeline#dbt#e2e testing#test scaffolding#metric testing#yaml fixture
One-click install
npx skills add https://github.com/constructorfabric/insight --skill metric-test-constructorfabric
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: metric-test
Source: https://github.com/constructorfabric/insight/tree/main/.claude/skills/metric-test
Command: npx skills add https://github.com/constructorfabric/insight --skill metric-test-constructorfabric

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Writing end-to-end tests for analytics metrics requires seeding bronze warehouse data, running dbt silver and gold models, and asserting API responses — a complex, error-prone process. This Skill scaffolds and validates the metric spec pair (a <name>.test.yaml fixture plus a test_<name>.py module) so the full bronze → silver → gold → analytics path is tested against a compose test-stand instance. ## Core Features & Use Cases - Spec scaffolding: Create a new <class>/<name>.test.yaml fixture with the required schemas and templates via /metric-test create <name> --metric <key> --tables <t1,t2>. - Offline validation: Resolve $ref references, schema-validate bronze records, and lint the pytest module without needing a running stand via /metric-test validate <path>. - Assertion guidance: Covers the assertion helpers (row/equals/contains, one/some over a view, approx), account bindings, identity aliases, duplicate-row dedup cases, and date-window boundary test design. - Use Case: When asked to "write a test for the emails-sent metric", the Skill produces the fixture seeding M365 bronze rows (including a re-sync duplicate) and a pytest module asserting collab.emails_sent period and peer values through POST /v1/metric-results. ## Quick Start Ask the assistant to write a test for a specific metric, for example: "Create a metric test for collab.emails_sent seeding m365 email activity rows."

Frequently Asked Questions about metric-test

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a test for an analytics metric?▼

Create a `<class>/<name>.test.yaml` fixture that seeds bronze tables with `$ref`-based records, plus a `test_<name>.py` module that calls POST /v1/metric-results and asserts rows with `r.row(key, view).equals(...)`. The `/metric-test create` command scaffolds both files.

How do I validate a metric test fixture without running the test stand?▼

Run the fixture loader with substitutions for tenant and supervisor placeholders, or use `/metric-test validate <path>`. It checks that every `$ref` resolves, each padded bronze record passes its JSON schema, and all placeholders are ones the run supplies.

How does the YAML fixture handle duplicate bronze records?▼

Two identical rows in the fixture simulate a real Airbyte re-sync duplicate, and the test asserts the metric dedups them rather than double-counting. After `$ref` resolution each row is padded to the full schema and validated with additionalProperties set to false.

Why does a seeded person drop out of team or department metrics?▼

Team attribution is a LEFT JOIN on lowercased email, so the person's M365 userPrincipalName must match their BambooHR workEmail case-insensitively. A mismatch resolves org_unit_id to NULL and silently excludes the person from the median and range computation.

What are the limitations of running the data-path suite?▼

The suite requires its own compose instance raised with `minimal`, since a stand seeded through silver is refused. Trees must be passed via `--tree=`, and linting with ruff must not run concurrently because the shared tests venv holds one dependency group at a time.

Related Skills

Explore More Software Engineering Skills→
databricks-solutionsdatabricks-solutions
5

04-grain-validation

Validate DataFrame grain against Gold layer DDL PRIMARY KEY definitions before merge operations.

Official
Intermediate
shafibabarshafibabar
1

go-mutation-test

Inject synthetic faults into Go source code to identify weak test assertions.

Community
Advanced
greenhatgreenhat

midenc-execution-trace

Trace MASM VM execution and print stack state after each instruction.

Community
Advanced
EyhenijEyhenij
1

testing

Enforces verifiability rules for unit and end-to-end tests across Vitest, Jest, and Playwright suites.

Community
Advanced