Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    data quality rules and monitoring

    1

    Testing and monitoring data quality manually is error-prone and slow.

    $2.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    data quality rules and monitoring

    data quality rules and monitoring

    Example session with this skill installed

    Run a quality check on the 'user_onboarding_v2' sample. I've attached the schema with not_null rules for email and unique constraints on user_id. Compare against the baseline for drift.

    • Read your context and instructions
    • Compiled the data quality rules

    The data quality execution for user_onboarding_v2 is complete. The run resulted in a needs_review status due to an alert-level failure rate in the uniqueness dimension.

    Data Quality Report - user_onboarding_v2, 2023-10-27

    1. Scope, sources and assumptions

    • Sources: E1 (CSV upload)
    • Dataset Approval: Approved by data_admin_01
    • Records: 1,000 supplied, 1,000 analysed, 0 excluded.
    • Assumptions: warn_rate defaults to 0.01 and alert_rate to 0.05 per user_supplied configuration.

    2. Quality ruleset

    RuleDimensionTypeSeverityWarnAlert
    R001Completenessnot_nullHigh0.000.01
    R002UniquenessuniqueCritical0.000.00

    3. Rule results and alerts

    RuleStatusEvaluatedFailingFailure rate
    R001ok100020.0020
    R002alert1000120.0120

    Overall pass rate: 98.60 percent.
    SLA Status: Breach detected on R002 (Uniqueness).

    4. Exception feed

    RuleRecordReasonObserved
    R002REC_089Duplicate IDuser_99
    R002REC_142Duplicate IDuser_99

    5. Trend dashboard

    No history was supplied.

    6. Distribution drift

    CheckColumnMetricValueWarnAlertStatus
    D001signup_sourcedist_shift0.080.050.10warn

    7. Remediation queue (draft)

    PriorityLevelSourceIdSeverityAction
    1BlockR002user_99CriticalManagement decision required

    8. Limitations and items for human review

    • Information gap — organizational validation required: No historical trend data available for this source.
    • [Computed result] The signup_source distribution has shifted significantly toward 'referral' compared to the baseline.

    Analytical estimate — validation required: The 1.2% duplication rate suggests a potential race condition in the upstream registration service.

    This analysis supports, and does not replace, the judgment of the accountable data owner.

    Next steps

    1. Review the remediation queue to resolve duplicate user_id records.
    2. Investigate the signup_source drift to confirm if marketing changes explain the shift.
    3. Upload historical run data to enable trend analysis.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Testing and monitoring data quality manually is error-prone and slow. This skill turns your schemas, business constraints, and record samples into an executable, deterministic quality ruleset with automated drift and trend detection.

    What it does

    • Rule execution runs completeness, validity, uniqueness, integrity, timeliness, and consistency checks.
    • Alerting & thresholds flags violations based on specific warn and alert rate tolerances you define.
    • Drift & trend analysis compares current samples against baselines and historical failure rates to find regressions.
    • Exception reporting generates a filtered feed of failing records and a prioritized draft remediation queue.
    • Governance logging tracks dataset approvals, source references (E-numbers), and consent-based exclusions.

    How it works

    1. Define the request by providing a dataset sample, schema, and quality expectations in a request.json format.
    2. Execute the validator using the scripts/run.py tool to perform exact arithmetic on the record sample.
    3. Review the envelope to check for SLA breaches, status warnings, or rules excluded due to access scope.
    4. Iterate on remediation using the draft queue to identify records requiring management intervention or correction.

    Frameworks & tools

    This skill uses a Python-based execution environment (python3) to run deterministic validation logic. It interfaces with JSON schemas for input/output structure and generates standard Markdown reports for data governance.

    Why this beats prompting it yourself

    Generic prompts often hallucinate statistics or fail at precision arithmetic across large record samples. This skill uses a dedicated script to ensure every failure rate, drift metric, and trend slope is calculated with exact logic rather than LLM estimation.

    Use cases

    • Validating a new batch of ETL output against referential integrity and format rules.
    • Monitoring for distribution drift in production datasets compared to a baseline.
    • Generating a prioritized exception feed for data stewards to review and remediate.
    • Tracking data quality trends across multiple runs to identify systemic pipeline issues.

    Known limitations

    This skill is for decision support only and cannot perform deletions or writes. It requires pre-approved datasets with source references and handles samples rather than full production databases.

    Decision support only; no direct data cleaning/deletion. Requires dataset approval and source references. Analysis is limited to the provided record sample.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 2 days ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables
    Listed2 days ago

    What's inside

    Frequently Asked Questions

    This skill uses a dedicated Python-based execution environment to perform exact arithmetic on your data, preventing the hallucinations and statistical errors common when asking a standard LLM to analyze record samples.

    The skill provides a deterministic validation script, a sample `request.json` template for defining your schemas, and documentation for integrating the Markdown governance reports into your existing workflow.

    It is designed to run in any Python 3 environment and can be integrated with AI agents that support script execution or custom tool calling via JSON-based inputs.

    No, this skill is strictly for decision support and read-only analysis; it generates remediation queues for stewards but lacks the permissions to delete or modify your production source data.

    The skill performs best with record samples and provided schemas; it is not intended to be plugged directly into full, multi-terabyte production databases without a pre-processing sampling step.