data-quality-assessment
Skill5 starsChanged 2 months agoMITView on GitHub
Systematic data quality evaluation covering completeness, accuracy, timeliness, consistency, and validity. Use when assessing data pipelines, reviewing data product quality, auditing data sources, defining quality SLAs, building data quality monitors, or when someone asks "is this data trustworthy?" or "how do we measure data quality?"
1 other project uses a version of this.
Adopted by
Most recently taken up by
- bingyue/fde-skillstheir own version, first seen Oct 13, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- bingyue/fde-skills★ 5same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository24 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- analystSkill
- arbitrage-audit-dataSkill
- dashboards-to-decisionsSkill
- databricks-data-product-reviewSkill
- data-consumer-discoverySkill
- data-model-designSkill
- data-pipeline-qualitySkill
- data-product-operatorSkill
- data-product-thinkingSkill
- data-product-validationSkill
- data-storytellingSkill
- data-team-operating-modelSkill
- data-team-positioningSkill
- duckdb-data-profilingSkill
- ethical-risk-assessmentSkill
- grill-me-dataSkill
- healthcare-cohort-designSkill
- healthcare-data-domainSkill
- healthcare-data-readiness-debriefSkill
- healthcare-data-reconciliationSkill
- metrics-definitionSkill
- research-synthesis-dataSkill
- stakeholder-alignmentSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 47
If data accuracy drops below threshold during development, pause deployment until the data lead investigates.
- Distinctive: line 53
If any stage fails, block promotion to the next stage
- Distinctive: line 69
Never ship a data product without at least completeness and accuracy checks in the pipeline.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Five Quality Dimensions
Score each dimension 1-5 when evaluating any data source or pipeline:
1. Completeness - What percentage of expected records and fields are present?
- Null rate per column
- Missing record detection (expected vs actual row counts)
- Required field coverage
- Score 5: <1% nulls in required fields. Score 1: >20% missing data.
2. Accuracy - Does the data reflect reality?
- Cross-field validation (age matches birth date, totals match line items)
- Reference data matching (codes exist in terminology tables)
- Statistical outlier detection
- Score 5: <0.1% error rate verified against gold standard. Score 1: Known systematic errors unresolved.
3. Timeliness - Is the data fresh enough for its intended use?
- Data freshness (time since last update vs SLA)
- Pipeline latency (ingestion to availability)
- Score 5: Real-time or within SLA. Score 1: Data is days/weeks stale.
4. Consistency - Does the same fact look the same everywhere?
- Format standardization (dates, codes, identifiers)
- Cross-system agreement (same patient, same record across sources)
- Naming convention compliance
- Score 5: Single source of truth, no conflicting definitions. Score 1: "Revenue" means 3 different things.
5. Validity - Does the data conform to business rules?
- Range checks (negative ages, future dates)
- Referential integrity (foreign keys resolve)
- Business rule compliance (diagnosis codes valid for encounter type)
- Score 5: All business rules enforced at ingestion. Score 1: Invalid data flows through unchecked.
Circuit Breaker Protocol
If data accuracy drops below threshold during development, pause deployment until the data lead investigates. Quality incidents in production destroy trust that takes months to rebuild.
Implement circuit breakers at every pipeline stage:
- Source validation (schema, nulls, ranges)
- Transformation validation (business rules applied correctly)
- Output validation (against reference data or prior period)
- If any stage fails, block promotion to the next stage
Quality vs Eval Distinction
Traditional quality checks: Is data complete? Does it match schema? Are calculations correct?
AI/ML evaluation (evals): Is the model telling the truth? Would a domain expert agree? Is it fair across user segments? Is it helpful or harmful?
Both are required for data products with AI components. Quality gates catch data issues. Evals catch model behavior issues. Neither substitutes for the other.
Practical Starting Point
Week 1: Audit the last 20 data quality issues. Write simple checks for the top 3 failure modes. Week 2: Label 100 production examples. Write your first automated quality monitor. Connect to CI/CD. Weeks 3-4: Monitor 10% of production data flow. Build the detect-fix-measure loop.
Never ship a data product without at least completeness and accuracy checks in the pipeline. Everything else can come later. Those two catch 80% of issues.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.