Use my Validio MCP and do data exploration. Help me identify the relevant datasets and critical fields so I know where I should start building validators. Then recommend in a table what sources and validators I should look at and why. Ask me for any business context needed to prioritize them.

Which datasets support [business process]? List the tables, which source they come from, and who owns them, then mark the 3 most business-critical sources I should look at.

Trace how [dataset/KPI/dashboard] is produced using available lineage and metadata. Identify its upstream inputs, documented transformations, and downstream uses. Highlight current gaps.

Profile [table/source] using the capabilities available to you. Summarize missing values, uniqueness, value distributions, and unusual patterns. Explain which findings matter most and which validators would cover them.

Give me a concise overview of my Validio workspace: monitored sources, existing validators, and recent incident activity. What are missing setups I need to create?

Where are the gaps in [source] monitoring? Check freshness, volume, schema, and field-level quality separately and tell me what is missing.

Which validators haven't produced an incident or a result in the last 30 days? Flag any that might be misconfigured or attached to a source that isn't polling.

Explain [source] to someone new to this dataset. Use its schema, descriptions, lineage, and validators to describe its purpose, dependencies, quality checks, and recent issues. Clearly separate documented facts from inferred meaning.

Review the Validio setup for [domain] and recommend three improvements we could make this week. Rank them by likely business value and implementation effort. For each, name the affected assets, supporting evidence, and a measurable success criterion.

Review incidents from the last 24 hours and any older incidents still open in [domain]. Rank the five that need attention first using severity, duration, and available downstream business context. Include incident references and the next action for each.

We need [dashboard/report] for [business decision] by [time and timezone]. Review relevant upstream incidents and monitoring freshness. Identify known issues that could compromise this deadline, what remains unverified, and what we should investigate first.

Review incidents in [pipeline/domain] over the last 48 hours. Identify groups that may share an upstream cause using lineage and timing. Explain the evidence for each relationship and which investigation could address the most downstream symptoms.

Which validators in [domain] generated the most incident groups in the last 30 days? Examine recurrence, duration, metric patterns, and available incident feedback. Identify candidates for investigation or tuning; do not assume frequent alerts are false positives.

Find the most persistent or recurring data quality issues in [domain] over the last 30 days. Rank them by observed disruption and downstream reach. Explain which deserve a permanent fix and identify recorded owners where available.

Investigate [incident ID or incident name] upstream through available lineage. Compare related incidents and their timing, accounting for validator windows. Identify the earliest observed anomaly and the most likely cause, distinguishing supporting evidence from hypotheses. Check whether values repeat across windows.

Trace the downstream dependencies of [incident ID/source]. Identify potentially affected tables, fields, and dashboards where available. Separate assets with observed quality issues from assets merely exposed through lineage, and prioritize what to validate first.

For [incident ID], compare the affected period with an appropriate healthy period using available metrics and profiling. Determine which fields or existing segments changed most, quantify the differences, and explain what additional data is needed to narrow the cause.

Investigate whether the anomaly on [source] is better explained by late data, missing records, duplicates, or a change in business activity. Use available validator metrics, configuration, and upstream evidence. Rank the explanations and propose a targeted check for each unresolved hypothesis.

Summarize the investigation of [incident ID]: timeline, affected assets, likely cause, supporting evidence, and unresolved questions. Recommend the next remediation and verification steps. Link the relevant Validio resources and distinguish work inside Validio from changes required in the data pipeline.

Based on our investigation of [incident ID], recommend the smallest set of validator changes that would detect this failure earlier. Compare with existing coverage and specify the source, metric, window, threshold approach, and any segmentation. Explain what each change would detect.

Review the schema, available profiling, and validators for [source], focusing on [critical fields] used in [business process]. Recommend missing checks for relevant risks such as nulls, duplicates, invalid values, freshness, and volume. Prioritize them and avoid duplicating existing coverage.

Review whether aggregate monitoring on [source] could hide issues within [country/customer/product]. Recommend suitable segmented validators, considering different segment patterns, sparse data, and the number of segments. Explain which additional failures they could reveal.

Review [validator] and its last 30 days of incidents alongside its windows, filters, and thresholds. Recommend configuration changes that could improve alert usefulness while preserving detection of important failures. Explain the trade-offs and how we should evaluate the changes.

Implement the validator changes we selected above for [source], checking for equivalent existing validators first. Verify the resulting configuration and report what was created or updated, what is active or awaiting approval, and when results can be assessed. Explain how we will confirm the intended failure is detected.