- Brand
- Document360
- Category
- Productivity
- Primary Subcategory
- Collaborative Docs & Workspaces
Integration details
Description
Document360 MCP helps authorized users search knowledge base content, inspect project versions and category trees, create and update articles or categories, publish or unpublish content, and move documents through workflow assignments from ChatGPT.
- Integration type
- Plugin
- Verification status
- Not applicable
- Platform
- ChatGPT
- Primary Subcategory
- Collaborative Docs & Workspaces
- Secondary Subcategories
- None listed
- Brand
- Document360
- Access
- Account required
- First tracked
- 2026-06-23
- Tool count
- 26
- Market
- US
The Primary Subcategory used for this profile’s headline score.
Other Subcategories where the Integration is listed.
Agent Discoverability
ChatGPT Plugin Discovery ScoreYour score is coming
ChatGPT now suggests Plugins on its own when they match a user's request.Your Plugin Discovery Score measures how often yours appears, and it will show here as soon as it’s ready.
What discovery looks like
Get alerts for Document360
Get updates when Document360’s Discoverability Score or category rank changes.
Competing in ChatGPT Collaborative Docs & Workspaces
View Category26 tools agents can invoke
Create a new article in a specified category and project version. Supports Markdown or Advanced WYSIWYG (HTML) content. The content format must match contentType. This tool does not publish articles. All drafts must be reviewed and published manually through Document360 portal. ## Use Cases - Adding a new article - Creating content in a category whose ID is known (e.g. from get-categories) ## Important Notes - First, identify the project version by calling document360-mcp-get-project-versions. If multiple versions exist, ask the user which one to use - do not assume. If only one version is available, proceed with it. - Next, fetch categories for that selected version using document360-mcp-get-categories with the projectVersionId. - When creating the article, ensure you pass the same projectVersionId along with a valid categoryId from that version. - For category selection: if there is one clearly relevant category, use it directly. If multiple relevant categories are found, ask the user to choose before proceeding. - Requires title, content, categoryId, and projectVersionId. Call document360-mcp-get-categories and document360-mcp-get-project-versions to obtain valid IDs. - For updates to existing articles, use document360-mcp-update-article instead. - Content format must match contentType: write only Markdown when contentType is Markdown; write full HTML with proper tags when contentType is Advanced WYSIWYG.
Create a new category in a specified project version. Can create a top-level category or a subcategory under a parent. ## Use Cases - Adding a new section or folder to the knowledge base - Creating a subcategory under an existing category (use parentCategoryId) ## Important Notes - First, identify the project version by calling document360-mcp-get-project-versions. If multiple versions exist, ask the user which one to use - do not assume. If only one version is available, proceed with it. - If creating a subcategory, fetch categories for that version using document360-mcp-get-categories with the projectVersionId to get a valid parentCategoryId. - For parent category selection: if there is one clearly relevant category, use it directly. If multiple relevant categories are found, ask the user to choose before proceeding. - Requires name and projectVersionId. - Omit parentCategoryId for a top-level category. - When categoryType = 1 (Page), the content parameter is required. For Markdown (contentType = 0), provide markdown text. For Advanced WYSIWYG (contentType = 2), provide HTML. - For updates to existing categories, use document360-mcp-update-category instead.
Add a new term to the project's business glossary. ## Use Cases - Recording the approved definition of a new product term - Adding a language-level term whose wording differs per language ## Prerequisites - Call document360-mcp-list-glossary-terms first: names must be unique per scope (global vs language-level). ## Important Notes - name is how articles refer to the term: the merge code {{glossary.<name>}} in Markdown content and data-glossary-name in HTML content. It cannot contain spaces. - The response's glossary.insertAs holds the exact text to insert: insertAs.markdown in Markdown content (contentType 0), insertAs.html in HTML content (WYSIWYG, contentType 1 or 2), where a bare merge code is shown to readers as literal text. - isGlobal = true (default): one term and definition for every language. langCode is optional; if given it must be a language of this project and is recorded as the term's language, as Customer API V3 does. - isGlobal = false: the term and definition are stored for langCode only (required). Add more languages afterwards with document360-mcp-update-glossary-term. - definition is HTML (plain text is accepted), up to 1000 characters of visible text. - Requires the Glossary feature on the project's plan and the content-reuse update permission.
Finds an exact or keyword text match across the documentation and returns a structured list of matching documents with their public URLs. Use this instead of document360-mcp-search when you are looking for a known literal string - a placeholder tag, a code string, an exact product term - because document360-mcp-search is semantic and will not reliably find one. By default it searches every project version the caller can access, but only ONE language - the current context language. To search other languages you must pass langCodes explicitly. ## Use Cases - Finding every document that contains a specific literal string, across the whole knowledge base - Producing a review or QA list scoped to a category, a version, or the entire knowledge base - Locating documents to read or edit with document360-mcp-get-article / document360-mcp-get-category - Auditing where a deprecated term, tag, URL or code sample still appears ## Prerequisites - Use document360-mcp-get-project-versions to get valid ids for versionIds. - Use document360-mcp-get-categories to get valid ids for categoryIds. ## Important Notes - Each match carries a type: 'article' or 'category'. A 'category' match is a PAGE CATEGORY - a category that holds its own page content. Read or edit it with document360-mcp-get-category and document360-mcp-update-category-content; document360-mcp-get-article cannot resolve its id. Folder and index categories hold no content and are never returned. - The match's id field is an article id when type is 'article' and a category id when type is 'category'. categoryId is always the CONTAINING category: an article's own category, or a page category's parent. - contentTypes narrows this to articles only or page categories only. Default: both. - matchType 'exact' requires the words to appear consecutively and in order, as a phrase, and stops the search from widening when the phrase is not found. matchType 'keyword' is ranked and unordered, so it returns more, looser results. - Matching is case-insensitive, and typo tolerance follows the project's search settings in BOTH modes - 'exact' does not mean zero typos. If a result looks like a near-miss rather than the literal string, check its snippet before concluding the term is present. - On a multi-language knowledge base, a match may exist in a language you did not search. If a term is not found, widen langCodes before concluding it is absent - use document360-mcp-get-project-versions to see which languages exist. An unknown langCodes value is rejected with the project's language list, so a typo cannot masquerade as 'not found'. - Indexed content is plain text with HTML removed, and is stored in segments of roughly 6 KB. A phrase that straddles a segment boundary is not found, and a match that exists only inside HTML markup cannot be found. - occurrenceCount counts the term inside the indexed content returned for that article, which is the best-matching segment - treat it as a lower bound, not a whole-article total. - Only the published, latest-draft and unpublished versions of an article are searchable. Older historical versions are not. - statuses defaults to published. Reader accounts always get published content only, whatever is requested. - categoryIds automatically includes every descendant category, up to 200 categories in total. A broader scope is rejected with an error rather than silently searching only part of it - scope to a deeper category, or omit categoryIds and filter the returned categoryId values yourself. - Paging past the first 100 results requires narrowing to a single version and language; scopedVersionIds and perVersionCounts in the response tell you how to split the work. - If notIndexedVersionIds is non-empty, those versions have no search index configured and returned nothing - that is not the same as having no matches.
Read an article using article ID or using an article URL. Returns article content and metadata. ## Use Cases - Reading full article content after finding an article via document360-mcp-search - Getting current content before updating: call get_article first, then update_article with the requested changes while retaining existing content ## Prerequisites - Requires either articleId or URL. ## Important Notes - Before updating an article, call this tool first to retain the existing content and change only the required parts. - The body is returned ONCE, in the articleContent object, keyed by its format: articleContent.markdown for a Markdown article, articleContent.html for every other editor. The key that does not apply is absent - the article object carries metadata only and no longer repeats the body. - articleContent.contentType names the editor: 'markdown', 'advancedWysiwyg', or 'wysiwyg' (legacy). When calling document360-mcp-update-article, pass contentType 0 for 'markdown' and 2 for 'advancedWysiwyg'; for 'wysiwyg' omit contentType entirely so the stored type is kept. - Send back the same format you received. Do not convert Markdown to HTML or the reverse - that rewrites the article in an editor it was not authored in.
Retrieves a list of all categories for a project version, including child categories and articles. Returns the full category tree; results are filtered by the current user's access permissions. ## Use Cases - Finding category IDs when creating an article (document360-mcp-create-article requires categoryId) - Browsing the knowledge base structure for a version - Verifying if a category exists and getting its path before create or update operations ## Prerequisites - Requires projectVersionId. Call document360-mcp-get-project-versions to get valid project version IDs.
Get a single category by ID, including its child categories, articles, and any category-level content. Access is validated against the current user's permissions. ## Use Cases - Fetching full details for one category (e.g. after listing with get-categories) - Verifying a categoryId before create-article or when moving an article - Reading category page content for Page or Index type categories ## Important Notes - categoryContent.content holds the page body ONCE, keyed by format: 'markdown' for a Markdown page, 'html' for every other editor. The key that does not apply is absent. - categoryContent.content.contentType names the editor: 'markdown', 'advancedWysiwyg', or 'wysiwyg' (legacy). Send the same format back to document360-mcp-update-category-content; do not convert between them.
List the tasks assigned to you in a workspace: workflow assignments, reader feedback and review reminders, sorted by due date. Same list as My tasks in the portal. ## Use Cases - Finding what you need to review or approve next - Checking overdue or due-this-week work ## Prerequisites - Get projectVersionId from document360-mcp-get-project-versions. ## Important Notes - taskTypes: any of "workflow", "feedback", "review_reminder". Omit for all. - due: "overdue", "today" or "this_week" (measured in your timezone via timezoneOffsetMinutes). - Includes tasks assigned to you directly or through a group. - Only the first 1000 tasks can be paged; narrow the filters for more. - Act on a task with its id: document360-mcp-snooze-task, document360-mcp-mark-task-done (workflow), document360-mcp-update-feedback-status (feedback), document360-mcp-set-article-review-reminder (review reminders).
List all project versions for the Document360 project. ## Use Cases - Getting version IDs for document360-mcp-search (versionFilter), document360-mcp-get-categories, document360-mcp-create-article, or document360-mcp-create-category - Verifying a version exists before creating or updating content - Discovering which versions are available
List the users and groups that can be assigned to a document's workflow. ## Use Cases - Resolving valid assignee IDs before calling document360-mcp-set-workflow-status with assignees ## Prerequisites - Requires documentId and projectVersionId. Set documentType to 'article' or 'category'. ## Important Notes - Use assigneeId values from assignableUsers for the assigneeIds parameter of document360-mcp-set-workflow-status. - Use assigneeId values from assignableGroups for the assigneeGroupIds parameter of document360-mcp-set-workflow-status; a group assignment covers all its members.
List the workflow statuses configured for the Document360 project. ## Use Cases - Discovering valid workflow status IDs before calling document360-mcp-set-workflow-status - Understanding the review pipeline (draft -> custom statuses -> published) ## Important Notes - Use the returned status id as workflowStatusId when setting a document's workflow status. - System statuses (initial draft, published, unpublished) are flagged via isSystemStatus.
Lists documentation pages as a structured, paginated set with their public URLs, filtered by scope and metadata rather than by a search term. A document is an article OR a page category - a category that holds its own page content. Both are returned by default. Use this when you need the contents of a category, a version or the whole knowledge base as a worklist, without searching for any particular text. ## Use Cases - Building a QA or review worklist for a category or an entire version - Finding everything changed within a date range - Listing draft or unpublished pages that still need work - Filtering by tag or label before reading pages with document360-mcp-get-article or document360-mcp-get-category ## Prerequisites - Use document360-mcp-get-project-versions to get valid ids for versionIds. - Use document360-mcp-get-categories to get valid ids for categoryIds. ## Important Notes - To search for specific text instead, use document360-mcp-find-text. - Each row carries a type: 'article' or 'category'. A 'category' row is a PAGE CATEGORY. Read or edit it with document360-mcp-get-category and document360-mcp-update-category-content; document360-mcp-get-article cannot resolve its id. Folder and index categories hold no content and are never listed. - The row's id field is an article id when type is 'article' and a category id when type is 'category'. categoryId is always the CONTAINING category: an article's own category, or a page category's parent. - contentTypes narrows this to articles only or page categories only. Default: both. - By default it lists from every project version the caller can access, but only ONE language - the current context language. Pass langCodes explicitly to include others; an unknown value is rejected with the project's language list. - statuses defaults to published. Reader accounts always get published content only, whatever is requested. - categoryIds automatically includes every descendant category, up to 200 categories in total. A broader scope is rejected with an error rather than silently searching only part of it. - tags and labels match on name, not id. - updatedFrom and updatedTo are ISO 8601 dates or date-times, treated as UTC, and both bounds are inclusive. - Paging past the first 100 results requires narrowing to a single version and language. - If notIndexedVersionIds is non-empty, those versions have no search index configured and returned nothing.
List the project's business glossary: each term's name, merge code, definition and per-language variants. ## Use Cases - Checking the approved definition of a product term before writing or reviewing an article - Finding a glossary entry's id before calling document360-mcp-update-glossary-term - Checking whether a term already exists before calling document360-mcp-create-glossary-term ## Important Notes - A glossary is either global (one term and definition for every language, isGlobal = true) or language-level (a term and definition per language, under languages). - scope mirrors the portal's glossary filters: "all" (default) lists global and language-level terms, "global" only global terms, "language" only language-level terms. - langCode lists language-level terms with content in that language; global terms are not included, so scope "global" cannot be combined with langCode. - Each language of a language-level term has translationStatus: "translated", "needs_translation", "in_progress" or "none" (the portal's per-language translation state). - Definitions are HTML, exactly as stored. - To use a term in an article or page, insert the insertAs form that matches the content's editor: insertAs.markdown (the merge code, e.g. {{glossary.sla}}) in Markdown content (contentType 0); insertAs.html (a <span data-glossary-name> element) in HTML content (WYSIWYG, contentType 1 or 2), where a bare merge code is shown to readers as literal text. The published page shows the current term, with its definition on hover. - Requires the Glossary feature on the project's plan and the content-reuse view permission.
List the articles and page categories in a workspace that carry a review reminder, earliest review date first. ## Use Cases - Finding content that is stale and needs review now - Planning upcoming reviews (scheduled reminders due in a date range) ## Prerequisites - Get projectVersionId from document360-mcp-get-project-versions. ## Important Notes - state "stale": flagged for review now. state "scheduled": fresh, becomes stale on reviewDate. Omit for both. - Archived content is excluded; only content you can access is listed. - Set or clear a reminder with document360-mcp-set-article-review-reminder.
Mark one of your workflow tasks done - the portal's Mark as done. ## Use Cases - Recording that you have reviewed or approved a document assigned to you in a workflow ## Prerequisites - Get the taskId from document360-mcp-get-my-tasks (taskTypes ["workflow"]). ## Important Notes - Records your approval only; the document's workflow status does not change. Move it with document360-mcp-set-workflow-status. - Works only when you have a pending approval on the task, directly or through a group. - Feedback tasks are closed with document360-mcp-update-feedback-status; review-reminder tasks by updating the article's review reminder with document360-mcp-set-article-review-reminder.
Publish an article - promotes its latest draft to the public/published version. ## Use Cases - Making a draft article live after creating or updating it ## Prerequisites - Requires articleId and projectVersionId. Use document360-mcp-search or document360-mcp-get-article to obtain the ID, and document360-mcp-get-project-versions for the version ID. ## Important Notes - Publishes the latest draft as-is; it does not modify content. Update the content first with document360-mcp-update-article if changes are needed. - An article in a read-only workflow status cannot be published, unless it is currently unpublished (re-publishing is always allowed). Check with document360-mcp-get-workflow-statuses and move it to an editable status with document360-mcp-set-workflow-status first.
Publish a Page category - promotes its latest draft content to the published version. ## Use Cases - Making a Page category's content live after creating or updating it ## Prerequisites - Requires categoryId and projectVersionId. The category must be a Page category (one that carries its own content). ## Important Notes - Only Page categories can be published with this tool. Folder and Index categories are not supported. - Publishes the latest draft as-is; update the content first with document360-mcp-update-category-content if changes are needed. - A category page in a read-only workflow status cannot be published, unless it is currently unpublished (re-publishing is always allowed). Check with document360-mcp-get-workflow-statuses and move it to an editable status with document360-mcp-set-workflow-status first.
Performs a semantic search to retrieve relevant article content for a query. Returns article content with metadata and content chunks. ## Use Cases - Content-based discovery: find articles and content chunks by meaning and context - Getting article IDs for use with document360-mcp-get-article or document360-mcp-update-article - Finding content that is conceptually related to the search query - To update an article: call this tool to find it, then document360-mcp-get-article to read it, then document360-mcp-update-article with the requested changes ## Important Notes - Chunk content is returned as Markdown. It is a retrieved excerpt, not the article source: to read or edit the whole article, take the articleId from metadata and call document360-mcp-get-article, which reports the article's real format. - This search is semantic. To find a known literal string - a placeholder tag, a code string, an exact product term - use document360-mcp-find-text instead; semantic retrieval will not reliably surface one. - To list documents by scope or metadata rather than by meaning, use document360-mcp-list-documents. - If versionFilter is not provided, search runs across all available project versions. - Use document360-mcp-get-project-versions to get valid version IDs for versionFilter.
Report a problem with a published article or Page category so the documentation team can fix it. ## Use Cases - The article did not answer the question, or left out a step (reason "missing") - The article contradicts the product or is out of date (reason "outdated") - The article is confusing ("unclear"), off-topic ("irrelevant") or has a broken link ("broken_link") ## Prerequisites - Pass exactly one of articleId or categoryId. categoryId must be a Page category (one that carries its own content). ## Important Notes - reason: "unclear", "outdated", "missing", "broken_link", "irrelevant" or "other". - comment is required for "outdated" and "other"; say exactly what is wrong (up to 2000 characters). - The feedback appears in the Feedback Manager as a "not helpful" vote from you, marked as sent by an AI agent. - Only published articles and Page categories you can read accept feedback, and not if feedback is switched off on them. - Sending the same feedback again within 24 hours is ignored (outcome "duplicate"). - Available to reader accounts too.
Set the workflow status of an article or a Page category, with optional due date, comment, and assignees. ## Use Cases - Moving a document through the review workflow (e.g. Draft -> In Review) - Assigning a workflow step to specific users with a due date ## Prerequisites - Requires documentId, documentType ('article' or 'category'), projectVersionId, and workflowStatusId. - Get a valid workflowStatusId from document360-mcp-get-workflow-statuses. - Get valid assigneeIds (users) and assigneeGroupIds (user groups) from document360-mcp-get-workflow-assignees. ## Important Notes - For categories, only Page categories support workflow status. Folder and Index categories are not supported. - The target status cannot be the published or unpublished system status. - Both individual users (assigneeIds) and user groups (assigneeGroupIds) can be assigned; a group assignment covers all its members. - When both assigneeIds and assigneeGroupIds are omitted: if the document already has assignees for the target status they are kept; otherwise the status's configured default assignees (from Workflow Designer) are applied automatically. Pass empty lists ([]) to force the document to be unassigned. - dueDate must be an ISO-8601 date (e.g. '2026-06-30') and not in the past.
Snooze one of your tasks until a date, or bring a snoozed task back. ## Use Cases - Deferring a task you cannot act on yet - Bringing a snoozed task back early ## Prerequisites - Get the taskId from document360-mcp-get-my-tasks (isSnoozed shows which are snoozed). ## Important Notes - until: ISO 8601 date in the future (e.g. "2026-11-02" or "2026-11-02T09:00:00Z"). Omit it to unsnooze. - A task is shared by its assignees, so snoozing hides it from all of them until the date or until the document changes. Only whoever snoozed it, or an Owner/Admin, can unsnooze it early. - Only the task's assignees (directly or through a group) and Owners/Admins can snooze it.
Unpublish an article - reverts the currently published version back to draft and removes it from the live knowledge base site. ## Use Cases - Taking a published article offline because it is outdated or needs rework ## Prerequisites - Requires articleId and projectVersionId. Use document360-mcp-search or document360-mcp-get-article to obtain the ID, and document360-mcp-get-project-versions for the version ID. ## Important Notes - Only a published article can be unpublished. A draft article cannot be unpublished. - An article that has newer draft changes in a later version cannot be unpublished while those draft changes are pending; publish or discard the draft first. - The content is retained; the article is simply no longer visible on the live site until it is published again with document360-mcp-publish-article. - An optional unpublishMessage may be provided (maximum 160 characters). - Unpublishing also moves the article to the project's unpublish workflow status, even when its current workflow status is read-only.
Unpublish a Page category - reverts its currently published content back to draft and removes it from the live knowledge base site. ## Use Cases - Taking a published Page category offline because its content is outdated or needs rework ## Prerequisites - Requires categoryId and projectVersionId. The category must be a Page category (one that carries its own content). ## Important Notes - Only Page categories can be unpublished with this tool. Folder and Index categories are not supported. - Only a published Page category can be unpublished. A draft category cannot be unpublished. - A category that has newer draft changes in a later version cannot be unpublished while those draft changes are pending; publish or discard the draft first. - The content is retained; the category is simply no longer visible on the live site until it is published again with document360-mcp-publish-category. - An optional unpublishMessage may be provided (maximum 160 characters). - Unpublishing also moves the category page to the project's unpublish workflow status, even when its current workflow status is read-only.
Update an existing article's title, content, category, visibility, or position. Call document360-mcp-get-article first, retain all existing content and formatting, and update only the part specified in the request. ## Content Handling Whether the article is Markdown or HTML: do not change any existing information or formatting unless explicitly requested. Update only the section or fields specified in the request; keep all other content and formatting intact. ## Use Cases - Editing article content after finding the article via document360-mcp-search and reading it with get_article ## Prerequisites - Requires articleId and projectVersionId. - Call get_article first. Retain existing content and formatting; update only the section or fields specified in the request. ## Important Notes - Do not change existing Markdown/HTML structure, styling, or other information unless that change is requested. - Do not alter variables or snippets. - Keep existing information intact; only update the specific part requested. - An article in a read-only workflow status cannot be updated. Check with document360-mcp-get-workflow-statuses and move it to an editable status with document360-mcp-set-workflow-status first.
Update a category's name, position, parent, or visibility. Does not update category page content; use document360-mcp-update-category-content for that. ## Use Cases - Renaming a category - Moving a category under a different parent (parentCategoryId) - Changing order or visibility (hidden) ## Prerequisites - categoryId and projectVersionId are required. ## Important Notes - For updating the content of a Page or Index category, use document360-mcp-update-category-content instead. - Changing visibility (hidden) also requires the publish permission, and it applies to the whole category subtree in the given language.
Update a category's page content and title. Only applicable to categories of type Page or Index. ## Important Notes - A category page in a read-only workflow status cannot be updated. Check with document360-mcp-get-workflow-statuses and move it to an editable status with document360-mcp-set-workflow-status first. - Call document360-mcp-get-category first and send back the SAME format it returned: markdown text when categoryContent.content.contentType is 'markdown', HTML when it is 'advancedWysiwyg' or 'wysiwyg'. Converting between the two rewrites the page in an editor it was not authored in. - The response echoes the saved body once, at categoryContent.content - the same path document360-mcp-get-category returns it at; the category object carries metadata only. That echoed categoryContent holds the body alone (no status or version numbers) - call document360-mcp-get-category for those.
Recommended guideHow to get your ChatGPT Plugin suggestedFollow the practical steps agents use to find, shortlist, and choose your Plugin.Read guideDocument360 ChatGPT Plugin FAQ
How the directory, categories and Discoverability Score work.
Read the methodologyHow do I improve Document360's ChatGPT Plugin discoverability?
The levers are the listing surface agents actually read: names, descriptions, keywords, tool metadata, and registry health. Which lever matters depends on where discovery breaks, which is what continuous measurement shows.
What are Document360 alternatives on ChatGPT?
As of 2026-11-02, Document360 competes with Abilitya Assistant, Accelflow Space, ArtifactBridge, AttachPage, Caipi, Clinked, Coda, Descry and 17 more in ChatGPT Collaborative Docs & Workspaces, ranked by public Discoverability Score.
Where is this profile measured?
This profile uses the geography attached to the latest public registry snapshot: US. Locale tags are intentionally omitted.