Office Agents is a monorepo of Microsoft Office Add-ins with integrated AI chat panels. Each add-in connects to major LLM providers using your own credentials (BYOK) and can read/write documents through built-in tools, a sandboxed shell, and a virtual filesystem.
Screen.Recording.2026-02-01.at.22.13.00.mov
| Package | Description |
|---|---|
@office-agents/sdk |
Headless SDK — agent runtime, tools, storage, VFS, skills, OAuth, web search |
@office-agents/core |
Chat UI (Svelte 5) — re-exports SDK + chat components, settings, sessions |
@office-agents/bridge |
Local HTTPS/WebSocket RPC bridge + CLI for inspecting a live Office add-in runtime |
@office-agents/excel |
Excel Add-in — spreadsheet tools, Office.js wrappers, system prompt |
@office-agents/powerpoint |
PowerPoint Add-in — slide/OOXML tools, Office.js wrappers, system prompt |
@office-agents/word |
Word Add-in — document text/structure/OOXML tools, Office.js wrappers, system prompt |
See each package's README for install instructions and tool documentation.
You can install skills from:
- a single
SKILL.mdfile, or - a folder that contains
SKILL.md.
Manage skills from the Settings tab.
- API key (BYOK): OpenAI, Anthropic, Google, Azure, OpenRouter, Groq, xAI, Cerebras, Mistral, etc.
- OAuth: Anthropic (Claude Pro/Max), OpenAI Codex (ChatGPT Plus/Pro)
- Custom endpoint: OpenAI-compatible APIs (Ollama, vLLM, LMStudio, ...)
In Settings you can configure:
- Provider, model, and auth method
- CORS proxy
- Thinking level
- Skills
- Web search/fetch providers and API keys
Configure web provider credentials in the Settings UI.
Supported providers:
- DuckDuckGo; search (free, but will rate limit easily)
- Brave; search
- Serper; search
- Exa; search, fetch
More often than not, basic fetch is good enough but requires a CORS proxy configured.
Note
This project is not production-ready and is not intended for publication to the Microsoft Add-in Store. Think of it as a framework and reference for building your own Office-based agents. That said, I'm happy to squash bugs — feel free to report them in Issues.
pnpm install # Install all dependencies
pnpm dev-server:excel # Start Excel dev server (https://localhost:3000)
pnpm dev-server:ppt # Start PowerPoint dev server (https://localhost:3001)
pnpm dev-server:word # Start Word dev server (https://localhost:3002)
pnpm start:excel # Launch Excel with add-in sideloaded
pnpm start:ppt # Launch PowerPoint with add-in sideloaded
pnpm start:word # Launch Word with add-in sideloaded
pnpm build # Build all packages
pnpm typecheck # TypeScript type checking (all packages)
pnpm lint # Run Biome linter
pnpm format # Format code with Biome
pnpm check # Typecheck + lint
pnpm validate # Validate Office manifestsDuring development, the Office taskpane auto-connects to a local bridge server. Use the bridge CLI to inspect the live add-in runtime, run tools, and manage VFS files:
pnpm bridge:serve # Start the bridge server (https://localhost:4017)
pnpm bridge:stop # Stop the bridge server
pnpm exec office-bridge list # List connected sessions
pnpm exec office-bridge inspect word # Inspect a session's tools & metadata
pnpm exec office-bridge tool word get_document_text # Run a tool against the live add-in
pnpm exec office-bridge screenshot word --out p.png # Screenshot a document page
pnpm exec office-bridge vfs ls word /home/user # Browse the VFSexec can run arbitrary JavaScript directly inside the Office taskpane — useful for debugging and closing the agentic dev loop
# Read the active Word document body text
pnpm exec office-bridge exec word --code \
"const body = context.document.body; body.load('text'); await context.sync(); return body.text;"
# Poke at taskpane globals / DOM
pnpm exec office-bridge exec word --code "return { href: window.location.href, title: document.title }"See packages/bridge/README.md for full CLI documentation.
MIT