Python-powered document generation for Microsoft Word

InkRider keeps Word as your editor and adds a runtime for the parts of the document that should be programmatic: anchored tables, charts, and computed text you can re-run, while prose and layout stay as easy as typing. Execute Python in the browser or on private servers; never treat the whole file as a build artifact.

Flexible & Secure Execution

Run Python entirely in your browser via WebAssembly (Pyodide/JupyterLite), or connect to your own private Jupyter Server for heavy workloads. Your data stays within your environment and never touches our servers.

Virtual File System (VFS) & DriveFS Bridge

Our advanced VFS bridge mounts your Word document as a local drive (/drive/doc/). Python can read and write internal assets directly using standard file I/O, powered by a PWA Service Worker XHR interception layer.

Strategic Storage: IndexedDB, Embedding & OneDrive

While local IndexedDB VFS allows storage to grow very large, it is subject to manual or policy-based browser cache erasure in Microsoft Edge WebView2. For permanent reliability, InkRider provides Document Embedding (CustomXML) for small assets and OneDrive Integration (/drive/cloud/) for large datasets.

Multi-Notebook Paradigm & Quarto Support

Tie specific, isolated notebooks (.ipynb), Python scripts (.py), and Quarto markdown files (.qmd) to different sections of your document. This prevents editing conflicts and enables true collaborative workflows.

Enterprise Security & Secrets Vault

A 4-layer defense system featuring SHA-256 cryptographic notebook signing, active XML integrity checks, and a zero-leakage Secrets Vault that injects API keys at runtime without saving them to the document.

Good Fit For

Documents with predictable content that can be generated by code.

Draft recurring reports or status updates with repeatable calculations and automated batch execution.

Prepare structured sections, styled grid-tables, and dynamic charts from lightweight data or cloud assets.

Mix short explanations, live markdown interpolation, and executable code in one Word authoring session.

Ensure strict compliance and data privacy with zero-backend code execution and local wheel mirroring.

InkRider is strongest when the document stays central and the code exists to support repeatable drafting, calculations, and structured output.

What you can do in the add-in

The core workflow stays simple: run code, generate output, and keep moving inside the document you are writing.

01

Anchor-Driven Document Insertion

InkRider's anchoring system is what makes outputs live inside the document rather than just appear in a side panel. Every cell can be tied to a specific position in your Word document; when the cell runs, its output is written directly into that location.

Create an anchor by clicking where you want the output in Word, which places a content control at the cursor and fills it either manually or automatically with the cell's output once it is ready

Re-run the cell at any time and the anchor updates in place, replacing only the managed region without touching surrounding text or formatting

Auto-insert mode writes every execution result to its anchor immediately after each run, keeping the document current without manual steps

Anchor links are stored inside the .docx file itself, so they survive saves, shares, and reopens across machines

Anchor the same cell to multiple document locations to repeat a computed value. For example, a summary figure can appear in both a header and a detail table and update together

Drift indicators badge affected cells when Word content at an anchor has been manually edited since the last insertion, preventing stale outputs from going unnoticed

02

In-browser Python execution & Watchdog

Run Python inside the add-in and keep calculations or content generation close to the document you are editing.

Runs Python directly in the browser, with no server setup required. The Pyodide runtime executes code in a sandboxed Web Worker; JupyterLite extends this with R and JavaScript kernel support

Proactive Resource Watchdog monitors WebAssembly heap memory and execution loops to prevent browser lockups

Offline-capable package mirror (local wheels) for strict corporate firewall environments without CDN dependencies

In-browser Secrets Management safely handles API keys and environment variables in local storage without document persistence

Local VFS Explorer allows managing large project workspaces in IndexedDB, with clear warnings regarding Edge browser cache erasure policies

03

Advanced Document Generation & Interpolation

Use Python and markdown together to generate rich, production-grade material placed directly into Word.

Interpolate live Python variables directly into your Markdown text using {{ expression }} syntax

Generate rich grid-tables with an advanced styling language supporting custom colors, alignment, and typography

Single-Button Re-Rendering instantly updates entire document sections or individual anchored outputs

Drift Indicators visually highlight when an inserted output has been manually modified in Word, preventing accidental overwriting

Cell Focus Mode isolates a single cell for distraction-free coding and debugging

04

Secrets Vault

API keys and credentials needed by Python code are stored in the Secrets Vault, a local-only store that is never written to the Word document, never exported, and never transmitted to any server.

Secrets are stored locally in the browser and injected into the kernel's environment (os.environ) at session start. Access them in code via os.environ.get('MY_KEY') with no changes to notebooks

Because secrets stay on your machine, the Word document can be freely shared or emailed, and recipients supply their own credentials through their own vault

Prevents the common mistake of hardcoding API keys into notebook cells that may later be shared, exported, or stored in SharePoint

Works with both the in-browser runtime and an external Jupyter Server. The vault injects into whichever session is active

Manage secrets from the Secrets tab in the Settings dialog: add, update, or delete individual entries, or clear the entire vault

05

Code Editor with IntelliSense

Write Python in a full Monaco editor (the same engine that powers VS Code) with completions that understand your project without a running kernel.

IntelliSense for the injected InkRider API, available before any cell is run

Completions for top-level functions and classes in your VFS Python scripts

Works in both JupyterLite and external Jupyter Server modes, with no kernel roundtrip

Syntax highlighting, bracket matching, find/replace, and line numbers out of the box

06

Collaborative, Batched & Cloud-Integrated

Designed for complex workflows, large teams, and enterprise asset management.

Execution Tool enables batched queue execution, running multiple notebooks sequentially or in parallel with auto-insertion

Document Embedding securely stores small assets and notebooks directly inside the .docx CustomXML structure for permanent portability

Direct integration with Microsoft OneDrive (/drive/cloud/) provides reliable, permanent cloud storage for large files and enterprise datasets

Built-in Acronym Generator automatically scans document text and code outputs to build and standardize terminology tables

Asset Locking prevents multiple users from overwriting the same code or document section simultaneously

07

Customizable Workspace & Built-in Tools

The InkRider panel system puts all relevant tools within reach without cluttering the editing area. Two configurable activity panels can be docked left, right, or at the bottom of the add-in; each panel independently switches between integrated mode (sharing space with the editor) and overlay mode (floating above it).

Tools panel: Execution Tool (batch queue runner), Session Manager (view and restart active kernels), VFS Explorer (browse and manage virtual filesystem files), and Cloud Storage (OneDrive browser)

Document panel: Anchors Viewer lists every anchor in the document with its status, drift state, last-run time, and a one-click jump to its position in Word

Logs panel: real-time log stream from the add-in runtime, filterable by level and component, useful for diagnosing unexpected execution or insertion behaviour

Focus mode isolates a single cell in a fullscreen view, which is useful for writing longer markdown sections or debugging complex code without panel distractions

Panel positions and visibility persist across sessions; each panel can be independently positioned, collapsed, or maximized

High-visibility anchor mode makes content controls visually prominent in Word so you can see exactly where each cell's output lands

Platform Support

InkRider is designed for current Microsoft 365 versions of Word on Windows, Mac, and the web.

Word for Windows

Versions

Microsoft 365 desktop

Notes

Designed for current Microsoft 365 desktop builds on Windows.

Fully supported

Word Online

Versions

Microsoft 365 online

Notes

Designed for Microsoft 365 online in modern browsers.

Fully supported

Word for Mac

Versions

Microsoft 365 desktop

Notes

Designed for current Microsoft 365 desktop builds on macOS.

Partially supported

Enterprise-Grade Security Architecture

Because InkRider embeds executable Python code inside Word documents, we built a 4-layer defense system to protect you from malicious macros and tampering.

XML Integrity Checking

InkRider actively monitors the underlying Word XML structure. If the document has been tampered with outside the add-in, execution is immediately halted.

Cryptographic Signatures

Every notebook payload is cryptographically signed with SHA-256 when saved. This prevents bad actors from unzipping the .docx file, altering the Python code, and executing it.

Input Sanitization

Strict JSON schema validation ensures that only safely formatted notebooks can be loaded into the Pyodide runtime.

Zero-Knowledge Data Privacy

Your code and data are processed entirely in your local browser or your privately hosted Jupyter Server. We have no backend servers reading your documents, guaranteeing total compliance.

A simple product model

Use InkRider in Word for document generation, then use the website for sign-in, billing, documentation, and support.

Run Python directly inside Microsoft Word

The browser-based runtime is built for content generation, calculations, and structured output in Word.

Supports add-in use and account tasks on the website

Sign in, manage billing, read the docs, and contact support on the website while the add-in stays focused on document work.

Stay in the Word workflow

InkRider is built so drafting, calculation, and insertion happen closer to the document you are writing.

Get help when you need it

Documentation covers the basics, and support is there for account, billing, and product issues.

Product at a glance

RuntimeJupyterLite
Main surfaceWord add-in
Account and billingCompanion site
Best fitDocument generation
Read the docs for setup and support details

Want to see whether InkRider fits your workflow?

Start with the documentation, check out the available plans, then reach out if you need product help, billing support, or a commercial conversation.