> For the complete documentation index, see [llms.txt](https://docs.blockbrain.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.blockbrain.ai/for-users/add-ins/word-add-in-quick-tour.md). # Word Add-in – Quick Tour A walkthrough of the Word Add-in: document summaries and clause lookup, rewriting selected text as tracked changes, native tables, dropped-in files and knowledge sources, model picker and chat history ## Overview The Blockbrain Word Add-in brings your AI workspace into Microsoft Word. With the document open you can **summarise it or ask about a specific clause, rewrite a selected passage as a tracked change, create a native table from a prompt, drop in files to compare against, switch between models and bots, and browse past conversations** — all without leaving Word. This guide covers the features visible in the current release: summarising and asking about the open document, rewriting selected text with suggestion cards and tracked changes, creating tables, working with dropped-in files and knowledge sources, selecting AI models and bots, and browsing your chat history. ## What the add-in can do {% columns %} {% column %} **Understand any document** {% embed url="" %} {% endcolumn %} {% column %} **Edit with confidence** {% embed url="" %} {% endcolumn %} {% endcolumns %} At a glance, these are the things you can do with the Word Add-in today: * **Summarising the open document** – full-document summaries, section summaries, and "what does the selected clause actually say" in plain language. * **Clause lookup and targeted Q\&A against the open file** – find the payment terms, the liability cap or the termination notice period without scrolling through the document yourself. * **Selection-scoped work** – the pane reads what you have highlighted (e.g. "Selection · 1448 characters · \~236 words") and confines the answer or the edit to exactly that passage. * **Rewriting a selected passage** – shorten, expand, simplify, or change tone and register; the most reliable way to write into the document. * **Suggestion cards with Apply, Apply with comment and Discard** – every change passes through a review gate; nothing reaches the document unreviewed. * **Edits land as real Word tracked changes** – visible in the Review tab and reversible with Reject All Changes, so an applied suggestion is always recoverable. * **Creating a native Word table from a prompt** – a real table object rather than pasted markdown, including turning prose into tabular form. * **Working with dropped-in files** – drag and drop files and folders into the pane, then ask questions or compare them against the open document. * **Attaching knowledge sources** – the + menu (Files, Emails, Library, Prompts, Skills) pulls your Blockbrain knowledge into the answer. * **Cross-lingual answering** – the reply follows the language of your prompt regardless of the document's language, with product and legal terminology preserved. * **Bot and model selection** – switch LLMs and assistants directly from the pane. * **Chat history** – prior conversations are reachable from the add-in and survive a pane reload; new chat, rename and delete are all available. * **Pane handling** – the pane is resizable and detachable and shows a token-usage indicator. * **Deployment** – installed centrally by IT via a manifest link in the Microsoft 365 admin center → Integrated Apps. ## What you'll need * Microsoft Word for Microsoft 365 (desktop or web) signed in with your work account. * The Blockbrain for Microsoft 365 Add-in installed and the user signed in. * A document open in Word that you want to discuss or edit with the AI. ## 1. Opening the add-in Once the add-in is installed, open any Word document and select **Blockbrain** in the Home ribbon (or via Add-ins). The Blockbrain pane opens on the right-hand side of the Word window and stays in sync with the document you are working on.
The pane can stay open while you scroll, edit or switch documents. It is resizable and can be detached into its own window, and a token-usage indicator at the top shows how much of the current conversation's context you have used. ## 2. Initiate all actions from the add-in side panel The right-hand pane is your home base for everything the add-in does. From here you can ask about or summarise the document, rewrite what you have selected, create a table, attach files and knowledge sources with the + menu, switch bots and models, and open your chat history. Whenever text is highlighted in the document, the pane shows the selection (e.g. "Selection · 1448 characters · \~236 words") and confines the answer or edit to exactly that passage.
## 3. Summarising and asking about the document With a document open, ask the assistant to summarise it (e.g. "Summarise this contract in five bullet points" or "What does section 7 say?"). You get full-document summaries, section summaries, and plain-language explanations of what a selected clause actually says — without leaving the pane.
### 3.1 Clause lookup and targeted Q\&A Instead of scrolling, ask for what you need: "Where are the payment terms?", "What is the liability cap?", "How long is the termination notice period?". The assistant finds the passage in the open file and answers with a reference to where it is.
### 3.2 Selection-scoped work Highlight a passage in the document and the pane picks it up (e.g. "Selection · 1448 characters · \~236 words"). Any question or edit you ask for is then confined to exactly that selection — useful for "explain this clause" or "is anything missing here?".
### 3.3 Cross-lingual answering The reply follows the language of your prompt, regardless of the language of the document. Ask in German about an English contract and you get a German answer, with product and legal terminology preserved as written. ## 4. Rewriting selected text with AI suggestions Highlight any text in the document and ask the assistant to shorten, expand, simplify, or change the tone or register. The assistant returns a suggestion card in the pane; nothing reaches the document until you apply it. This is the most reliable way to write into the document. ### 4.1 Generate a suggestion Select the paragraph or phrase you want the AI to work on. In the side panel, type your instruction (e.g. "Tighten this paragraph" or "Rewrite for a non-technical audience") and submit. The panel displays the proposed change next to the original. ### 4.2 Review the suggestion card The suggestion card shows what will be inserted, replaced or removed next to the original. Every card offers three choices: **Apply** writes the change into the document, **Apply with comment** does the same and attaches a note, and **Discard** drops the suggestion. The review gate is the point of the feature — nothing reaches the document unreviewed.
### 4.3 Apply with a tracked comment If you want to leave a note alongside the change — for reviewers, or as a reminder to yourself — choose Apply with comment. A small dialog opens for your comment text; the comment is attached to the modified passage so it shows up in the Word review pane. Choose Apply if you only want the edit and not the annotation, or Discard to drop the suggestion. Both Apply options write the change immediately.
### 4.4 Edits land as tracked changes Applied suggestions are written as real Word tracked changes, not plain overwrites. They are visible in Word's Review tab like any other revision, and a bad apply is always recoverable: use Reject All Changes (or reject a single change) to return to the original text.
### 4.5 Track what has been applied Conversations survive a reload of the pane, so you can pick up exactly where you left off. From the list you can also rename a chat or delete it. ## 5. Creating a native Word table from a prompt Ask for a table in plain language (e.g. "Turn the delivery milestones in section 4 into a table with date, deliverable and owner") and the assistant inserts a real Word table object at the cursor — not pasted markdown. This also works for converting prose into tabular form, and the result can be formatted and edited like any other Word table.
## 6. Choosing an AI model and bot Different tasks suit different models. The selector at the top of the pane lists all LLMs that have been approved for use with the selected bot, and the Agent tab lets you switch to any of the approved bots on your system. **Auto Router** picks a suitable model for each request automatically and is a good default if you are unsure.
**Tip:** pick a fast, lightweight model (Haiku, Flash) for short rewrites and a more capable model (Opus, Sonnet, GPT 5) for long-form drafting or analysis. Auto Router is a good default if you are unsure.
## 7. Chat history Open the chat history from the pane to see your previous conversations with Blockbrain — both document-specific and general chats. Each entry shows the chat title, and selecting one reopens that conversation alongside the current document. Use New Chat at the top of the list to start a fresh thread. The search box helps you find a specific past chat by keyword from its title or contents. Conversations survive a reload of the pane, so you can pick up exactly where you left off. From the list you can also rename a chat or delete it.
## 8. Working with files and knowledge sources You can give the assistant additional context beyond the open document — either by dropping in files from your computer or by attaching sources from your Blockbrain knowledge. ### 8.1 Dropping files and folders into the pane Drag files or whole folders from your computer straight into the pane, or use the upload action to pick them (PDFs, slides, spreadsheets, other Word documents, images, and so on).
Once added, the assistant can reference the files in the conversation, summarise them, compare them against the document you have open, or quote passages from them. *\[Screenshot placeholder: files dropped into the pane and a comparison question about them]* ### 8.2 Attaching knowledge sources with the + menu The **+** menu next to the prompt box connects the conversation to your Blockbrain knowledge: **Files**, **Emails**, **Library**, **Prompts** and **Skills**. Whatever you attach here is resolved and fed into the answer, so you can, for example, check a contract against your internal guidelines or reuse a saved prompt.
## 9. Pane handling The pane is resizable — drag its edge to make room for longer answers — and can be detached into its own window so it sits next to Word rather than inside it. A token-usage indicator at the top shows how much of the current conversation's context you have used, so you know when to start a new chat.
## 10. Recommended working patterns ### Drafting and refining Write your first draft, then walk through it section by section. Highlight a paragraph, ask for a rewrite, and either accept or refine the suggestion. Use Apply with comment when collaborating so reviewers can see why a passage changed. ### Document Q\&A For complex documents (policy papers, contracts, reports), ask the pane instead of searching manually: where a clause is, what it says in plain language, and whether anything is missing. Highlight a passage first to keep the answer scoped to it. ### Cross-document research Drop supporting files or a whole folder into the pane so the assistant can reason across the open document and the attachments. This is useful for comparison, citation hunting and gap analysis. ### Building tables from prose When a section lists milestones, prices or responsibilities in running text, ask for a table instead of building one by hand. The result is a native Word table you can keep formatting yourself. ## Troubleshooting and tips * **Panel not visible:** Re-open the add-in from Word's Home ribbon → Add-ins → Blockbrain for Microsoft 365. The pane docks on the right by default. * **No model response:** Check the model selector — if a model is rate-limited, switch to another one or select Auto Router. * **Suggestion did not apply:** Make sure the original selection still exists in the document. If the text was edited after the suggestion was generated, regenerate the suggestion against the new selection. * **Lost a chat:** Use the search box in the chat history to look up a previous chat by keyword from its title or contents.

--- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.blockbrain.ai/for-users/add-ins/word-add-in-quick-tour.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.