SkillVaultskills Browse all 500 skills

Data · Version 1.4.0 · Reviewed 2026-08-02

Warehouse Model Reviewer

Make data systems more correct and operable for grain and fan-out review and slowly changing dimension design with evidence, explicit trade-offs, and a verification plan.

4 method steps 6 documented failure modes 5 diagnostic checks 7 quality gates

Reviews dimensional and semantic-layer models for grain correctness, fan-out risk, slowly changing dimensions, and metric consistency.

₹149 one-time

Get this skill archive

What this skill helps you do

  • Grain and fan-out review
  • Slowly changing dimension design
  • Metric definition consistency

How Warehouse Model Reviewer works

You provide

Model definitions, declared grain, and conflicting metrics

It inspects

Joins that fan out rows and violate the declared grain

It decides

Grain corrections and a single metric definition

You verify

The same metric returns one number across dashboards

What it checks first

Warehouse Model Reviewer reviews dimensional and semantic-layer models for grain correctness, fan-out risk, slowly changing dimensions, and metric consistency. Use it when the work involves Grain and fan-out review, Slowly changing dimension design, Metric definition consistency.

  1. Whether the failure is systematic across a class of inputs or random, which separates a capability gap from a sampling issue.
  2. Whether evaluation data overlaps training or prompt-development data, which invalidates the measurement.
  3. Token distribution of inputs and outputs, since cost and latency are driven by the tail, not the mean.
  4. Whether the system has a defined behavior for low confidence, or always produces an answer.
  5. Version pinning across model, prompt, retrieval, and tools, because an unpinned component makes regressions unattributable.

Failure modes it recognizes

  • Silent quality regression after a provider updates a model behind an unversioned alias.
  • Evaluation overfitting where the prompt was tuned on the same examples used to score it.
  • Cost and latency dominated by a small number of very long inputs that were never in the test set.
  • Tool-calling loops where the model retries a failing tool without a bounded attempt budget.
  • Confident fabrication when context is insufficient because no refusal path was defined.
  • Distribution shift where production inputs diverge from the evaluation set over time.

Answers it will reject

  • Judging quality by reading a few outputs, which cannot detect a regression of a few percent.
  • Using a larger model to fix a problem caused by missing context, paying more for the same failure.
  • Fine-tuning before exhausting prompting and retrieval, which is slower to iterate and harder to reverse.
  • Using an LLM judge without validating the judge against human labels on the same rubric.

Decision rules it applies

  • Establish a labeled evaluation set and a baseline before changing anything; without a baseline there is no improvement, only change.
  • Pin every version and change one component at a time.
  • Define and test the refusal path explicitly; a system that cannot say "I do not know" will fabricate.
  • Budget latency and cost on p95 token counts, not averages.

Evidence it asks for

  • Score per input class (easy, hard, adversarial, no-answer) so aggregate scores cannot hide a broken class.
  • Log model version, prompt version, and retrieval version on every request for regression attribution.
  • Track p50 and p95 tokens and cost per successful task, not per call.

The method inside

  1. Map the artifact, actors, boundaries, and invariants relevant to grain and fan-out review.
  2. Trace concrete failure or abuse paths for slowly changing dimension design; do not report checklist items without a mechanism.
  3. Prioritize metric definition consistency findings by impact, likelihood, confidence, and cost of correction.
  4. Recommend the smallest defensible change, then define how an independent reviewer can verify it.

Deliverables

  • Grain and fan-out review assessment
  • Slowly changing dimension design decision and action plan
  • Metric definition consistency verification checklist

Evidence requirements

  • Schema, access patterns, query plans, or event contracts
  • Volume, cardinality, retention, and freshness
  • Consistency, latency, and migration constraints

Quality gates

  • Every material claim traces to supplied evidence or is labeled as a hypothesis.
  • The response follows the declared deliverable contract.
  • No execution, access, measurement, or verification is invented.
  • Secrets and personal data are redacted rather than repeated.
  • The user receives a concrete independent verification step.
  • The relevant failure modes in this domain were considered rather than only the reported symptom.
  • No listed anti-pattern was recommended as a solution.

Example task

Input

Our revenue metric returns different totals depending on which dashboard it is used in.

Expected output

Different totals from one metric almost always mean a grain mismatch: the fact table is being joined to a dimension that fans out rows, inflating the sum in some contexts. Identify the declared grain of each model and find the join that violates it before changing the metric definition...

Boundaries and compatibility

Ideal for

  • Grain and fan-out review: produce a decision or artifact grounded in supplied evidence.
  • Slowly changing dimension design: produce a decision or artifact grounded in supplied evidence.
  • Metric definition consistency: produce a decision or artifact grounded in supplied evidence.

Out of scope

  • Recommending indexes without a workload
  • Treating eventual consistency as universally acceptable

Agent compatibility

  • GitHub Copilot custom agents
  • Claude Agent Skills / SKILL.md
  • Any instruction-following chat model

Tool policy: Advisory by default. No tools are assumed. If the host provides tools, use read-only evidence gathering unless the user explicitly approves a scoped write or execution action.