CtrlK
BlogDocsLog inGet started
Tessl Logo

testland/data-quality-conventions

Reference catalog of data-quality conventions - when to choose dbt-tests vs Great Expectations vs Soda, column-level vs table-level coverage, severity tiering, SLA and freshness conventions, and common anti-patterns to avoid. Use when designing coverage for a new data product or auditing an existing one.

77

Quality

97%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Overview
Quality
Evals
Security
Files

Quality

Content

92%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-crafted, concise, and highly actionable conventions catalog with a clear design sequence and embedded feedback loops. The one weak spot is progressive disclosure: everything lives inline in SKILL.md with no reference files, so the top-level is not kept lean as an overview.

Suggestions

Split the detailed tables (engine-selection matrix, starter-check set, anti-patterns, retirement criteria) into reference files under references/ and keep SKILL.md as a concise overview with one-level-deep links, so the top-level stays lean.

Move the 'Common anti-patterns' and 'When to retire a check' sections into a references/anti-patterns.md, clearly signaled from the overview, to reduce inline density.

Add a short 'Quick start / decision flow' section near the top that links to each detailed section, improving navigation and making the progressive-disclosure structure explicit.

DimensionReasoningScore

Conciseness

Lean and dense with decision tables and concrete conventions; it does not explain concepts Claude already knows and every 'Why' column earns its tokens. Not the level below because there is no padding or unnecessary explanation to tighten.

3 / 3

Actionability

Provides concrete, copy-paste-ready guidance - specific check names, 'freshness < 2d', severity tier behaviors, naming examples like 'not_null_orders_email', and 'meta:' owner blocks. Absence of executable code is not penalized for this instruction-only catalog, per the scoring notes.

3 / 3

Workflow Clarity

Presents a clear design sequence (engine selection -> column checks -> table checks -> severity -> freshness -> naming -> retirement) with embedded feedback loops such as check-health retirement criteria and 'land new checks at warn for two cycles, then promote'. Not a destructive/batch context requiring validation gates.

3 / 3

Progressive Disclosure

The ~120-line SKILL.md holds the entire catalog inline with no bundle or reference files, so it is the full content rather than an overview pointing to detailed materials. The 'under 50 lines with no need for external references' exemption does not apply here, so well-organized sections alone do not reach 3. Not the level below because the body is well-sectioned with a clear intro and no nested/deep references.

2 / 3

Total

11

/

12

Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description that names concrete capabilities, uses natural trigger terms, explicitly covers both what and when, and cleanly distinguishes itself from related engine-specific skills. No first/second person voice is present.

DimensionReasoningScore

Specificity

Lists multiple specific concrete capabilities - engine choice (dbt-tests vs Great Expectations vs Soda), column- vs table-level coverage, severity tiering, SLA/freshness conventions, and anti-patterns - rather than vague language.

3 / 3

Completeness

Explicitly answers both what it is ('Reference catalog of data-quality conventions...') and when to use it ('Use when designing coverage for a new data product or auditing an existing one.').

3 / 3

Trigger Term Quality

Uses natural domain terms a user would actually say - 'dbt-tests', 'Great Expectations', 'Soda', 'data product', 'auditing an existing one' - giving good coverage of trigger phrases.

3 / 3

Distinctiveness Conflict Risk

Carves a clear niche distinct from sibling engine skills ('those tell you the how of running checks; this tells you which checks and where'), making wrong-skill triggering unlikely.

3 / 3

Total

12

/

12

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Reviewed

Table of Contents