MCP App Store
by KeenEthics
Workiva icon

Workiva MCP Server for ChatGPT

By Workiva

Overview

Connect ChatGPT to the Workiva Platform for Workiva accounts hosted in Workiva US Production. This release is available to supported users in the United States, Canada, and Mexico. Use the Workiva connection to help users find and work with resources available through their Workiva account.

How to connect

  1. ChatGPT: Open the Workiva listing in the ChatGPT app directory and choose Connect. ChatGPT listing Workiva exposes 70 tools here.

Tools

documents__get_cell_range

ChatGPT
Read sparse cells for an A1 range from a direct document table reference.

documents__get_chart

ChatGPT
Get a chart's type, formatting, and configuration (title, legend, axes, series styling). Pass the ID of a chart element returned by get_section or get_floating_content. The chart holds no data values; read them by passing the returned tableId to get_cell_range. SeriesOrientation says whether each series is a row or a column of that table. ChartType specifies which of three base types the chart is: combo (line, column, bar, area, waterfall, etc.), pie (pie & doughnut), and scatter (scatter, bubble, advanced bubble).

documents__get_floating_content

ChatGPT
Get floating content (shapes, text boxes, images, charts, tables, lines) for a section. Pass floatingContentId and revision returned by get_section.

documents__get_footnotes

ChatGPT
Get one or more footnotes from a document as an ordered page of rich-text paragraphs. Requires a revision from the tool call where you got the footnoteIDs. When nextCursor is returned, call again with the same documentId, footnote IDs, revision, and the returned nextCursor.

documents__get_image

ChatGPT
Get an image from a document so you can view it. Pass the documentId and the ID of an image element returned by get_section or get_floating_content. Returns the image as a PNG or JPEG, downscaled if it is large.

documents__get_section

ChatGPT
Get a section from a document's outline. Returns the section name and a page of body paragraphs; when incomplete is true, call again with the returned revision and nextOffset to read the rest.

documents__list_document

ChatGPT
List documents in a workspace, optionally filtered by name. Results are sorted by name. Returns document IDs and names; pass a document ID to list_section to see its sections, then call get_section with a document ID and a section ID to read one.

documents__list_section

ChatGPT
List the sections in a document's outline. Returns section IDs and names only, no content.

documents__search_fulltext_document

ChatGPT
Search the full text content of documents. Use this when you know a phrase or wording inside a document but not its name; use list_document instead to find documents by name. Returns matching documents, each with excerpts of the matched text. Pass a match's documentId to list_section, or a match's documentId together with a snippet's sectionId to get_section to read the section the excerpt came from. Snippets without a sectionId matched the document name rather than its content and cannot be used with get_section. A hasMore of true means more documents matched than were returned; narrow the search term or raise the limit.

grc_graph_tools__grc_graph_data_types

ChatGPT
List every vertex type known to this workspace's graph structure. Call this first when you need the exact vertex_type string required by grc_graph_get_form_list, grc_graph_get_form_definition, or grc_graph_get_reviews_and_logs, and you only have a natural-language reference (e.g. 'controls', a typo, or a human-facing label) rather than the machine vertex type. Takes no arguments — always returns every non-hidden vertex type for the current workspace, under dataTypes (keyed by vertex type). Each entry has: label (human-readable name), properties[] (name, label, isRequired), and edges (keyed by edge label, each with target vertex type, label, single, required). Where the type has a form definition, properties/edges are limited to the fields that appear on that form; otherwise every structure field is returned. Property names and edge labels can be passed as properties/edges to grc_graph_get_form_list. An edge's target is the vertex type on the other end: use that type's properties as the edge's property_names, or list its instances with grc_graph_get_form_list.

grc_graph_tools__grc_graph_get_form

ChatGPT
Retrieve a hydrated form for a graph vertex, with current property values and edge data filled in. Returns: title (entity display name), _properties (resolved scalar values keyed by property name), _relationships (edges to related entities keyed by form field name, each with label, direction, count, targetType when known, and targets as {id, display}; edges with nothing to show are normally left out, and an entry with an empty targets list means the edge is present and empty; _relationshipsTruncated means the server cut the result, so counts and targets may be incomplete), _edges (table-style edge results keyed by query hash), _references, permission, and section summary. An edge missing from _relationships is either empty or not on this form; use grc_graph_get_form_definition to see which fields the form has. IMPORTANT: _references and _relationships targets contain vertex IDs pointing to related entities. NEVER display these IDs to the user; show the display name. To fetch a related entity's details, it is a different type than this form's entity, so first call grc_graph_get_form_definition with vertex_type set to the relationship's targetType, then call this tool with that form_definition_id and the ID as root_id. Use grc_graph_get_form_definition to discover the formDefinitionId first. The root_id is a vertex ID from other GRC tools (e.g. a control ID from list_controls, a risk ID from list_risks, a program ID from grc_graph_get_sox_programs). Returns ResourceNotFound if the definition or root vertex does not exist.

grc_graph_tools__grc_graph_get_form_definition

ChatGPT
Retrieve the form definition (template/schema) for a vertex type. This returns the form structure WITHOUT data — use grc_graph_get_form with the returned formID to get a hydrated form with actual values. Provide vertex_type or form_definition_id (the server validates the combination). Approximate vertex_type names (e.g. 'controls') are resolved automatically; if no type matches, suggestions are returned instead. Returns a Form object with: formID (pass to grc_graph_get_form), vertexType, displayName, intent (audit/sox/issue/resource/system), and items[] — an ordered list of FormItem entries (sections with nested inputs, tables, edge lists, notes, components). Returns ResourceNotFound if the vertex type has no form definition.

grc_graph_tools__grc_graph_get_form_list

ChatGPT
List all form instances for a vertex type. Returns a list of entries, each with: display (human-readable instance name), rootId (vertex ID — pass to grc_graph_get_form to fetch the full form), and permission. Approximate type names (e.g. 'controls', 'audit program') are resolved automatically; if no type matches, suggestions are returned instead. Optionally request extra per-instance data: properties, edges (with properties of the connected entities), attachment counts, and suggested change counts. PREFER ONE CALL OVER MANY: to answer questions that filter, compare, or summarize instances (e.g. 'controls owned by Alice', 'which issues are open'), call grc_graph_data_types with the vertex_type to get field names, then request the needed properties/edges here instead of calling grc_graph_get_form for each instance. Edges from grc_graph_data_types list a target type; use that type's properties as the edge's property_names (e.g. owner -> Person -> ['name']). Unrecognized field names are reported under fieldWarnings with suggestions. Returns an empty list if no instances exist.

grc_graph_tools__grc_graph_get_phase_type_statuses

ChatGPT
Retrieve the set of possible statuses for test phases (e.g., not started, in testing, in review, complete). Use these to interpret the status field on test phase data.

grc_graph_tools__grc_graph_get_prior_toc_metadata

ChatGPT
Retrieve metadata from the prior testing period for a test of control. Returns the previous audit program's ID, name, and associated spreadsheet reference.

grc_graph_tools__grc_graph_get_report

ChatGPT
Execute a report and return its row data. Pass the dataSourceId from grc_graph_get_reports as resource_id. Returns resolved column headers and cell data with display names instead of raw IDs. Includes truncated (true when more rows exist beyond max_rows) and recordCount. Results may be paged: when the response includes a cursor, pass it back as the cursor argument to fetch the next page. Some reports execute asynchronously and may return a PROCESSING status while this tool polls for completion; if it is still processing when the tool call times out, retry the call later. Pass refresh=true to force the server to bypass any cached result. If notFound is true the report does not exist. For user-centric reports (isUserCentric=true in the report list), pass view_as_person_id='*' to include all users. Set include_views_only=true to list available views without executing the report. Pass view_id to execute a specific view instead of the default.

grc_graph_tools__grc_graph_get_reports

ChatGPT
List all reports in the workspace as a folder tree. Returns a root ReportFolder with nested folders[] and reports[]. Each report has: dataSourceId (pass to grc_graph_get_report), reportTitle, reportDescription, isUserCentric, and staticReportMetadata (isStatic, lastRefreshed, refreshedByName). Navigate folders by name to find a specific report.

grc_graph_tools__grc_graph_get_reviews_and_logs

ChatGPT
Retrieve reviews (approvals, rejections, assignments with reviewer and status) and a chronological activity log for a form. Identify the form by form_definition_id or vertex_type, plus an optional form_instance_id.

grc_graph_tools__grc_graph_get_sox_programs

ChatGPT
List SOX audit programs for the current workspace. Returns each program's vertexId (use as audit_program_id or program_id in other SOX tools), name, and archived status. Start here to discover programs before drilling into tests of control. Do not display raw vertex IDs to the user.

grc_graph_tools__grc_graph_get_test_form

ChatGPT
Look up a single test of control by ID and retrieve its detailed form data, including the list of its test phases (with vertex IDs), attribute matrices, timing, form section IDs, and control information. For a program-level overview use grc_graph_get_test_forms_by_program_id instead.

grc_graph_tools__grc_graph_get_test_form_list

ChatGPT
Retrieve a tabular summary of test forms, queryable by program_id or specific toc_ids. Returns column-oriented data suitable for listing and comparison, including phase-level scheduling fields such as test phase due dates. Optionally filter by phase_type_id. For full detail on one TOC use grc_graph_get_test_form.

grc_graph_tools__grc_graph_get_test_forms_by_program_id

ChatGPT
Retrieve all tests of control for an audit program and the program's test form groups (vertexId, name, accessibility). Returns program-level data including form groups and group assignments. Optionally filter by phase_type_id. Do not display raw vertex IDs to the user. To show full details for a test form, call grc_graph_get_form with the vertex ID as root_id.

grc_graph_tools__grc_graph_get_test_phase_form

ChatGPT
Retrieve detailed form data for a single test phase (phase vertex IDs come from grc_graph_get_test_form), including attribute matrices, timing window, testing status (not started / in testing / returned / in review / review complete), supplemental files, and section form IDs.

grc_graph_tools__grc_graph_get_test_phase_review_plan_items

ChatGPT
Retrieve the review plan for a test phase: who reviews, in what role, their status, due date, and which review is currently active. Each item maps a reviewer to a review type with an order and status.

grc_graph_tools__grc_graph_get_test_phase_review_plan_types

ChatGPT
List all review plan type definitions (e.g., preparer, reviewer, approver). Each type has a vertexId and reviewType name. Use these to interpret the reviewType on review plan items.

grc_graph_tools__grc_graph_get_test_phase_types

ChatGPT
Retrieve test phase type definitions (e.g., design effectiveness, operating effectiveness) for an audit program or test of control. Use the returned type IDs to filter other tools by phase_type_id.

grc_graph_tools__grc_graph_get_test_steps

ChatGPT
Get test steps and their attributes for a test phase. Each step has an ID, label, description, ordered position, and attribute values. Identify the phase by SOX IDs (test_of_control_id, optionally narrowed by test_phase_id) or form instance IDs (root_id + form_definition_id).

grc_tools__grc_aggregate_controls

ChatGPT
Aggregates controls in the current workspace by various dimensions — process, owner, significance, location, system, dimensionality type, and test/assessment aging. Returns summary counts for dashboards and overview reports. A control is counted under each of its owners, so owner buckets can add up to more than controlTotalCount. Definitions have no owners, so they appear only in the null owner bucket. Generally, filter to one dimensionalityType (DEFINITION, INSTANCE, or STANDALONE); without that filter the counts mix definitions with their instances. When filtering to instances (dimensionalityType = INSTANCE or using a definitionId filter), the result __typename is ControlsInstanceAggregate and includes a drift field with field-override metrics; otherwise __typename is ControlsBaseAggregate and drift is absent. Do NOT use to list individual controls — use grc_list_controls. Terminology: Multi-instance controls (MIC): a Definition is a template shared across locations/systems; an Instance is one concrete copy of a Definition scoped to a specific location and/or system; Standalone is the standard single-entity model — one control record, not split into definition/instance pairs, used when MIC is not enabled. Standalone controls have locations/systems (arrays); Instances have location/system (singular); Definitions have neither. Use grc_get_governance_workspace_settings to determine if MIC is enabled.

grc_tools__grc_aggregate_process_risks

ChatGPT
Aggregates process risks in the current workspace by dimension — process, sub-process, owner, and mitigating-control coverage. Returns summary counts for dashboards and overview reports. Do NOT use to list individual risks — use grc_list_process_risks.

grc_tools__grc_get_audit_area

ChatGPT
Loads a single audit area by its ID with full details including area type, inherent and residual risk scores (a residual that is missing or has no score has not been calculated or set, or the methodology defines none; an unscored area returns no scores), owner, related process risks, last audited project, and current user permissions. Use when the user references a specific audit area by ID and needs comprehensive information. Do NOT use to enumerate or search audit areas — use grc_list_audit_areas.

grc_tools__grc_get_business_response

ChatGPT
Returns the business (management) response recorded on a remediation issue, including its survey answers, for one issue by ID. Use when you need the business response for an issue. Requires the issue ID.

grc_tools__grc_get_control

ChatGPT
Loads a single control by its ID. The response includes a __typename field (ControlStandalone, ControlDefinition, or ControlInstance) indicating the control type. Standalone and Instance controls include full details: test plans, reference assets, active issues, active assessments, and testing signal. Definition controls are templates — they include instance count and risks, but do not have their own test plans, issues, assessments, or testing signal (those belong to their instances). Use when the user references a specific control by ID and needs comprehensive information. Do NOT use to enumerate or search controls — use grc_list_controls. Terminology: Multi-instance controls (MIC): a Definition is a template shared across locations/systems; an Instance is one concrete copy of a Definition scoped to a specific location and/or system; Standalone is the standard single-entity model — one control record, not split into definition/instance pairs, used when MIC is not enabled. Standalone controls have locations/systems (arrays); Instances have location/system (singular); Definitions have neither. Use grc_get_governance_workspace_settings to determine if MIC is enabled.

grc_tools__grc_get_flowchart_data

ChatGPT
Retrieves a single flowchart by ID with its full contents: all diagram cells (types, positions, connections, styling) plus the associated processes, sub-processes, and embedded controls and process risks. Use when the user references a specific flowchart and needs its diagram or associations. Requires the flowchart ID. To get narrative or deeper details for a process, use grc_get_process with the process ID.

grc_tools__grc_get_governance_workspace_settings

ChatGPT
Loads the governance workspace settings: whether multi-instance controls (MIC) are enabled, and who is emailed when a control suggestion is submitted. Use to check workspace configuration before interpreting control data — when MIC is enabled, controls may be Definitions or Instances rather than Standalone. When controls.multiInstance.enabled is null or false, MIC is not enabled and all controls are Standalone. controls.suggestions.subscribers lists the users notified about new control suggestions; it is notification routing only and confers no permission to triage them. Returns a single settings object with no required inputs.

grc_tools__grc_get_issue

ChatGPT
Loads a single remediation issue by its ID with full details including findings, sources (workpaper, audit project, assessment, or CPM task), associated controls/processes, custom fields, and resolution metadata. Use when the user references a specific issue by ID and needs comprehensive information about it. Do NOT use to enumerate or search issues — use grc_list_issues.

grc_tools__grc_get_policy

ChatGPT
Loads a single policy by its ID, including its latest document and sections. Use when the user references a specific policy by ID and needs its current content. Read each section's plainText field for its readable content, with the section title as the first line; use content only if you need the underlying ProseMirror/TipTap document JSON structure. The returned document is the newest version and may be an unpublished draft — check status before presenting content as the effective policy. Do NOT use to enumerate or search policies by metadata — use grc_list_policies; to find policies by topic or content, use grc_search_policy_sections.

grc_tools__grc_get_process

ChatGPT
Loads a single process by its ID with full details including associated sub-processes, controls, risks, flowchart, and narrative. Use when the user references a specific process by ID and needs comprehensive information about it. Do NOT use to enumerate or search processes — use grc_list_processes.

grc_tools__grc_get_process_risks_by_ids

ChatGPT
Loads multiple process risks by their IDs in a single call. Use when the user has explicit process risk IDs and needs to fetch them all together. Do NOT use to search or filter process risks — use grc_list_process_risks. IDs that do not exist are silently omitted from the result.

grc_tools__grc_get_sub_process

ChatGPT
Loads a single sub-process by its ID with full details including parent process, controls, risks, flowchart, and narrative. Requires both the sub-process ID and its parent process ID — use grc_list_sub_processes first if the parent ID is unknown. Do NOT use to enumerate sub-processes — use grc_list_sub_processes.

grc_tools__grc_list_analytics_views

ChatGPT
Get the available GRC analytics views. Call this FIRST to discover available view names, measures, dimensions, segments, time dimensions, and filters before building a query. Field names returned here are the exact strings to use in grc_run_analytics_query. Each view corresponds to one dataset; pick the view that matches what the user is asking about.

grc_tools__grc_list_audit_area_types

ChatGPT
Lists all audit area types defined in the current workspace. Use to discover available type values for filtering grc_list_audit_areas.

grc_tools__grc_list_audit_areas

ChatGPT
Lists audit areas in the current workspace with filtering, search, and pagination. Returns each area's type, owner, and inherent and residual risk scores; a residual that is missing or has no score has not been calculated or set, or the methodology defines none, and an unscored area returns no scores. Use to enumerate audit areas or find areas matching criteria. Do NOT use to fetch a single audit area by known ID — use grc_get_audit_area.

grc_tools__grc_list_controls

ChatGPT
Lists controls in the current workspace with filtering, search, and pagination. Each node includes a __typename field (ControlStandalone, ControlDefinition, or ControlInstance) indicating the control type — use this to access type-specific fields. The orderBy parameter accepts 'createdAt'. Use to enumerate controls or find controls matching criteria. Do NOT use to fetch a single control by known ID — use grc_get_control. Terminology: Multi-instance controls (MIC): a Definition is a template shared across locations/systems; an Instance is one concrete copy of a Definition scoped to a specific location and/or system; Standalone is the standard single-entity model — one control record, not split into definition/instance pairs, used when MIC is not enabled. Standalone controls have locations/systems (arrays); Instances have location/system (singular); Definitions have neither. Use grc_get_governance_workspace_settings to determine if MIC is enabled.

grc_tools__grc_list_issue_actions

ChatGPT
Lists the remediation actions (tasks) on an issue by ID, with pagination, including each action's schedule and status. Use to enumerate an issue's action items. Requires the issue ID. This is read-only; it does not create or modify actions.

grc_tools__grc_list_issues

ChatGPT
Lists remediation issues in the current workspace with filtering and pagination. Use to enumerate issues, find open issues by severity, status, or type, or look up issues attached to a specific control, process, or risk. Supports filtering by type (AUDIT_FINDING, CONTROL_ISSUE, GENERAL_ISSUE). Do NOT use to fetch a single issue by known ID — use grc_get_issue. Returns a paginated IssueConnection.

grc_tools__grc_list_locations

ChatGPT
Lists all locations defined in the current workspace with optional filtering and pagination. Use to enumerate or look up location IDs for use when filtering controls or risks. Returns a paginated LocationConnection with id and name.

grc_tools__grc_list_policies

ChatGPT
Lists policies in the current workspace with filtering and pagination. Each result includes identifying fields (name, document type, catalog status) plus its department and category, if assigned. searchTerm matches policy name or description, not document content — for topic- or content-based search, use grc_search_policy_sections. Do NOT use to read a policy's document content — use grc_get_policy. catalogStatus is the catalog-level lifecycle (NEW/PUBLISHED/RETIRED); it does not indicate whether the latest document version is published — check grc_get_policy's document status for that.

grc_tools__grc_list_process_risks

ChatGPT
Lists process risks in the current workspace with filtering, search, and pagination. Use to enumerate process risks or find process risks matching criteria. Do NOT use to fetch specific process risks by known IDs — use grc_get_process_risks_by_ids. Returns a paginated ProcessRiskConnection.

grc_tools__grc_list_processes

ChatGPT
Lists processes in the current workspace with filtering and pagination. Use to enumerate processes or find processes matching criteria. Do NOT use to fetch a single process by known ID — use grc_get_process. Returns a paginated ProcessConnection.

grc_tools__grc_list_sub_processes

ChatGPT
Lists sub-processes in the current workspace with filtering and pagination. Use to enumerate sub-processes or find sub-processes matching criteria. Do NOT use to fetch a single sub-process by known ID — use grc_get_sub_process. Returns a paginated SubProcessConnection.

grc_tools__grc_list_systems

ChatGPT
Lists all systems defined in the current workspace with optional filtering and pagination. Use to enumerate or look up system IDs for use when filtering controls. Returns a paginated SystemConnection with id and name.

grc_tools__grc_run_analytics_query

ChatGPT
Execute an analytics query against a view discovered via grc_list_analytics_views and return rows of data. Use grc_list_analytics_views first to discover available field names. Field names must be in 'view_name.field_name' format (e.g. 'view_controls_unified.count'). When presenting results to the user, speak in business terms only — never mention field names, segment names, tool names, or query parameters. IF THIS TOOL RETURNS AN ERROR: do NOT resend the same query — an identical query will fail again. Instead, simplify it. Start from a minimal query (one measure plus one segment), confirm it returns data, then add detail back one piece at a time; when a query fails, remove 'filters' first, then reduce 'dimensions'. To narrow to a specific value, prefer adding that field to 'dimensions' and reading the relevant row(s) from the result. Retry an unchanged query at most once, and only if you suspect a momentary glitch.

grc_tools__grc_search_policy_sections

ChatGPT
Searches policy document sections by semantic similarity to a natural-language query. This is a similarity search, not exact keyword or substring matching — differently worded content can still match, and an empty result does not prove that no relevant policy exists. Results are not restricted to published policies unless you pass documentStatuses — each result's documentStatus may be an unpublished draft; check it before presenting matched text as the effective policy. Use when the user describes a topic or concept and needs relevant policy content surfaced, not when they already know which policy they want. Do NOT use to look up a policy by its known ID (use grc_get_policy) or to enumerate/filter policies by metadata (use grc_list_policies).

list_workspaces

ChatGPT
List workspaces the authenticated user belongs to (organization and workspace id/name). Optional filter in the filter argument: use organizationId eq <id> or workspaceId eq <id> for exact match; for display names use infix contains, e.g. workspaceDisplayName contains test or organizationDisplayName contains 'Acme' (quote literals with spaces). Combine clauses with and (case-insensitive).

public_api_documentation__list_pages

ChatGPT
List all available documentation pages. Use read_page to retrieve the full content of a page.

public_api_documentation__read_page

ChatGPT
Read the full content of a documentation page by its relative path (e.g. '2026-01-01/authentication.md'). Do NOT pass full URIs — only the relative path portion.

public_api_documentation__search_docs

ChatGPT
Search the API documentation by keyword. Returns matching pages with context snippets. Use read_page to retrieve the full content of a result.

spreadsheets__get_cell_range

ChatGPT
Read sparse cells for a sheet-local A1 range through the sheet's backing table.

spreadsheets__list_sheet

ChatGPT
List the sheets (tabs) in a spreadsheet. Returns sheet IDs, names, and table references for follow-on get_cell_range reads.

spreadsheets__list_spreadsheet

ChatGPT
List spreadsheets in a workspace, optionally filtered by name. Results are sorted by name. Returns spreadsheet IDs and names that can be passed to list_sheet or get_cell_range.

wdata__download_query_result

ChatGPT
Downloads a completed query result as CSV text. The result must have status COMPLETE (check with get_query_result). Paginated by byte range (default 100KB per page). For large results, consider adding a LIMIT clause to the query.

wdata__get_query

ChatGPT
Returns a saved query's SQL text, parameters — including valid choices for select-mode parameters — and table dependencies. Use search_query first to find the query ID, then call this before run_query to see what parameters the query accepts and what values are valid for them.

wdata__get_query_result

ChatGPT
Retrieves metadata for a query result. Provide queryResultId to look up a specific result (e.g. one returned by run_query), or queryId to fetch the latest COMPLETE result for that query without re-running it. Returns status (RUNNING, COMPLETE, ERROR, CANCELLED, NOT_STARTED), columns, row count, duration, bytes scanned, and any error message. When status is ERROR, the error field contains the specific Presto failure reason (e.g. 'line 1:8: Column ''foo'' cannot be resolved' or 'Table ''x'' does not exist') — use it to diagnose and correct the SQL before retrying with run_query. Keep polling every 1-2 seconds while status is RUNNING.

wdata__get_table

ChatGPT
Returns detailed schema for a single wdata table by ID, including all columns with their types and modes, plus description, table type, status, and timestamps.

wdata__list_table

ChatGPT
Lists available wdata tables in a workspace. Returns each table's id, name, type (DATA, DIMENSION, HIERARCHY), status, description, columnCount, and timestamps. Paginated. Set includeSchema=true to also return a columnNames array of non-system column names per table — useful for discovering which tables have relevant columns without calling get_table on every candidate. Note: best used with a narrow filter — large workspaces can return many columns. Call get_table on the 1-2 identified tables for full column detail before writing a query. Optionally narrow results with the filter parameter (oData-style); see its description for supported expressions and examples.

wdata__run_query

ChatGPT
Executes a query and returns immediately with a queryResultId and RUNNING status. Provide either queryId (an existing saved query, with optional parameter overrides — call get_query first to discover its parameters) or queryText (ad-hoc SQL; Presto dialect, SELECT only) — exactly one of the two. Prefer queryId over queryText: saved queries are pre-validated and parameterized; only write ad-hoc SQL if no saved query covers the request. When a saved query accepts parameters, prefer overriding them to filter the result set rather than running unfiltered and filtering after download — this reduces bytes scanned and rows returned. Poll get_query_result every 3 seconds (up to 40 attempts); if status is ERROR or CANCELLED, report the error message to the user and stop. On COMPLETE, proceed to download_query_result. FROM clause formats by source type: wdata tables (default): FROM "workspaceId"."tableId"; Workiva spreadsheets: FROM spreadsheets."workbookId"."sheetId"; saved queries: FROM queries."workspaceId"."queryId"; external Snowflake tables: FROM external."workspaceId"."tableId"; PCAF datasets: FROM pcaf."workspaceId"."sourceProviderId"; business data: FROM business_data."workspaceId"."tableId"; ESG programs: FROM program."workspaceId"."tableId" or FROM program_v2."workspaceId"."tableId"; program dimensions: FROM program_dimension_v1."workspaceId"."tableId"; program tags: FROM program_tag_v1."workspaceId"."tableId"; Common Object Model: FROM com."workspaceId"."tableId"; identity services: FROM iam."workspaceId"."tableId"; graph reports: FROM graph."workspaceId"."tableId"; frameworks explorer: FROM explorer."workspaceId"."tableId" or FROM explorer_v2."workspaceId"."tableId".

wdata__search_query

ChatGPT
Searches saved queries in a workspace by matching a search term against query names. Omit searchTerm to list all saved queries. Returns each query's id, name, description, and parameterCount. A parameterCount above 0 means the query accepts parameters — call get_query on that query to see their names, types, and valid choices before running it. Paginated.

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/workiva"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/workiva)