quality · datawithbaraa — Claude Code skill for data quality reports
quality — a Claude Code skill that reports data checks and whether tables are safe to use. Give it a date or period to inspect failures, warnings, quarantined rows, and freshness.
MIT ★ 106 updated 2026-11-06
/plugin marketplace add DataWithBaraa/claude-code-data-engineering-skills
/plugin install data-engineering@data-with-baraaWhat quality does
The skill reads run records, check results, quarantine data, and source reconciliation when a project has one. It reports passed, warned, or failed checks, affected rows, and tasks skipped by each run. With no period supplied, it uses the last month.
For each table, the report compares freshness with its service level and assigns a verdict: safe to use, use with care, or do not use. The result is a single HTML file based on references/quality.html. It writes only that page; it does not change checks, rerun jobs, or release quarantined rows.
What it helps with
- identify checks that failed or raised warnings
- see which files and rows were quarantined
- compare table freshness with its service level
- decide whether a table is safe to use today
- compare record counts and totals with the source
Example prompts for the agent
- Report data quality for yesterday and mark which tables are safe to use.
- Create a quality report for the last month.
- Show quarantined rows and table freshness for this week.
How to install quality
1Plugin marketplace
/plugin marketplace add DataWithBaraa/claude-code-data-engineering-skills/plugin install data-engineering@data-with-baraaThis installs the whole data-engineering plugin with all its skills
2Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill quality -a claude-code3Manually: put the files in place
~/.claude/skills/quality/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/quality/
1Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill quality -a codex2Manually: put the files in place
.agents/skills/quality/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill quality -a cursor2Manually: put the files in place
.cursor/skills/quality/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install DataWithBaraa/claude-code-data-engineering-skills skills/quality2Manually: put the files in place
.github/skills/quality/SKILL.md1Skills CLI (Vercel)
npx skills add DataWithBaraa/claude-code-data-engineering-skills --skill quality -a gemini-cli2Platform CLI
gemini skills install https://github.com/DataWithBaraa/claude-code-data-engineering-skills --path skills/qualityManually: put the files in place
.devin/skills/quality/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about quality
How do I install quality in Claude Code?
Run /plugin marketplace add DataWithBaraa/claude-code-data-engineering-skills, then /plugin install data-engineering@data-with-baraa.
What does the quality skill do?
It gathers run, check, quarantine, freshness, and source reconciliation results into an HTML report with a verdict for each table. It does not alter checks or rerun jobs.
What format does quality use for its report?
It creates one HTML file with period metrics, daily check results, quarantine details, and stale tables.
Source text
SKILL.md · MIT
--- name: quality description: Use to see which checks failed or warned, what was quarantined, how fresh the tables are, and whether each table is safe to use today. argument-hint: "[date or period]" --- # Quality The job: say what the checks stopped and why, and whether each table can be trusted today. It reads, it never changes a check. Input: a date or a period, the last month by default. ## Steps 1. The status of the runs: succeeded or failed, the check that failed a run, and the tasks that were skipped. 2. Every check, layer by layer: passed, warned or failed, with the table and the rows it affects. 3. The quarantine: every stopped file and row, with the rule it broke. 4. Freshness: when every table the consumers read was last rebuilt, against its service level. 5. The reconciliation with the source, when the project has one: the counts and the totals on both sides. 6. The checks that fail most, and the ones that warn every day and nobody looks at anymore. 7. The checks that never failed: when each one was last tested with bad data, so you know it still works. 8. The verdict per table: safe to use, use with care, or do not use, with the reason. 9. Give the result as one web page. ## Output One HTML file, like [references/quality.html](references/quality.html): the numbers of the period, the verdict per table, when the data was ready, every check on every day, the checks that failed most, the quarantine per day and the stale tables. Read it for the layout before you write the page. ## Reads - the run records, the check results, the quarantine, the reconciliation ## Writes - the web page only ## Never - Never changes a check, never reruns a job, never releases a quarantined row.
© Data With Baraa · MIT · DataWithBaraa/claude-code-data-engineering-skills
Similar skills
All skills for Data and databases →quality by other authors comes first.