Repository navigation
Document that Ask AI uses personal warehouse credentials, and its expiry message (+8 more) - #1640
Merged
Merged
Conversation
Contributor
|
Preview deployment for your docs. Learn more about Mintlify Previews.
|
Contributor
Contributor
Documentation validation✅ All blocking checks passed. Component system0 finding(s) across the corpus. Findings on lines this pull request touched are annotated inline in Files changed. |
jpetey75
reviewed
Oct 30, 2026
| </Note> | ||
|
|
||
| <Note> | ||
| Row functions operate across your entire results table, with no equivalent to SQL's `PARTITION BY`. If a dimension in your rows splits the data into categories (for example a region or product column), `offset`, `index`, `offset_list`, `lookup`, and `row` don't reset at each category — an `offset(${orders.total_revenue}, -1)` can return a value from a different category's row. If you need an offset or lag that resets per group, either: |
Contributor
There was a problem hiding this comment.
this shouldnt be a callout
Collaborator
Author
|
Pushed Un-callout the offset() partitioning explanation on row functions for the review comments above.
Comment again to change more, or merge when it is right. |
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Docs changes Cloudy's docs writer made from support threads on Friday 30 October 2026. One commit per change, described below in order; review by commit and comment on any to drop or change. Every commit builds on current main, so merge whenever it is green, including midday for an urgent fix: later changes that day start a new PR.
Note that a provider admin can block the MCP OAuth callback URL
Pages
What changes
Why
Readers connecting an OAuth MCP server (e.g. Atlassian Rovo) can hit a provider consent screen blocking the Lightdash callback URL, and the docs gave no explanation or next step for that provider-side block.
Checked against
packages/backend/src/controllers/aiAgentsController.tsCloudy found this docs gap while working a support thread. Started from this internal Slack thread (Lightdash workspace only).
Document composer file attachments (CSV and document)
Pages
What changes
Why
No page explained what files the AI agent composer can attach, their formats, size limits, or the drag-and-drop fallback, even though two other pages assumed readers already knew.
Checked against
packages/frontend/src/ee/features/aiCopilot/components/ChatElements/AgentChatInput.tsxpackages/common/src/ee/AiAgent/threadFileTypes.tspackages/frontend/src/ee/features/aiCopilot/hooks/useThreadFileAttachment.tspackages/frontend/src/ee/features/aiCopilot/hooks/useCsvSourceAttachment.tsA teammate answered a support question that the docs should have covered. Started from this internal Slack thread (Lightdash workspace only).
Document how to attach a file to a single agent prompt
Pages
What changes
Why
The agent docs never explained how to attach a file to a single prompt, so readers only found knowledge-document upload (.md/.txt) and assumed Excel or CSV prompt attachments weren't possible or required External sources.
Checked against
packages/common/src/ee/AiAgent/threadFileTypes.tspackages/frontend/src/ee/features/aiCopilot/components/ChatElements/AgentChatInput.tsxpackages/frontend/src/ee/features/aiCopilot/hooks/useThreadFileAttachment.tsCHANGELOG.mdCloudy found this docs gap while working a support thread. Started from this internal Slack thread (Lightdash workspace only).
Document locked dashboard filters caused by hiding their field
Pages
What changes
hiddenin Dimension configuration and Metric configuration to that explanationWhy
Readers only knew locked filter chips as editor-applied locks, so a chip locked because its field was hidden looked unexplained and un-fixable.
Checked against
packages/frontend/src/features/dashboardFilters/lockedFilters.tspackages/frontend/src/features/dashboardFilters/LockedFilter.tsxpackages/common/src/utils/filters.tspackages/common/src/utils/i18n/uiStrings.tsCloudy found this docs gap while working a support thread. Started from this internal Slack thread (Lightdash workspace only).
Link the upgrade-check section to the upgrade-preflight agent skill
Pages
What changes
upgrade-preflightskill as an automated alternative to the manual CLI stepsWhy
The upgrading page described the manual upgrade-check steps but never pointed readers to the upgrade-preflight agent skill that automates them, even though the skills page already links the other way.
Checked against
skills/upgrade-preflight/SKILL.mdskills/lightdash-release-review/SKILL.mdCloudy found this docs gap while working a support thread. Started from this internal Slack thread (Lightdash workspace only).
Note that row functions don't partition by group
Pages
What changes
offset,index, etc. run over the whole table, noPARTITION BY-equivalentPARTITION BYwindow functions or pivot functions for a per-group offset/lagWhy
The row-functions reference only warns that sort order affects results, so readers using
offseton a results table with a category dimension in the rows get silently wrong period-over-period values instead of a pointer to the per-group alternatives.Checked against
packages/common/src/utils/tableCalculationFunctions.test.tspackages/common/src/utils/tableCalculationFunctions.tsA teammate answered a support question that the docs should have covered. Started from this internal Slack thread (Lightdash workspace only).
Warn that offset() spans the whole results table, not per dimension
Pages
What changes
PARTITION BYand to Period comparison for within-group offsetsWhy
Readers used offset() for year-over-year comparisons on results with more than one dimension and got silently wrong values, because the page didn't say the function ignores dimension grouping or point to the partitioned alternatives.
Checked against
packages/common/src/utils/tableCalculationFunctions.tsCloudy found this docs gap while working a support thread. Started from this internal Slack thread (Lightdash workspace only).
Un-callout the offset() partitioning explanation on row functions
Pages
What changes
PARTITION BYexplanation is now a plain paragraph, not a<Note>calloutWhy
A reviewer asked that this explanation not be styled as a callout.
A reviewer's comments on this pull request asked for this change. Started from this internal Slack thread (Lightdash workspace only).
Document that Ask AI uses personal warehouse credentials, and its expiry message
Pages
What changes
Why
Readers couldn't find that personal credential projects make Ask AI use those credentials, or that an expired sign-in makes Ask AI reply with misleading text instead of the reconnect prompt charts and Explore show.
Checked against
packages/backend/src/ee/services/ai/utils/toolErrorHandler.tspackages/common/src/types/errors.tspackages/frontend/src/hooks/toaster/ApiErrorDisplay.tsxpackages/backend/src/services/WarehouseClientFactory/personalCredentialOverlay.tsCloudy found this docs gap while working a support thread. Started from this internal Slack thread (Lightdash workspace only).