office-automation

OmegaUse-OfficeVal is a benchmark for evaluating LLM agents on long-horizon office-suite tasks with economic grounding, comparing human costs and LLM inference costs. It includes 100 tasks requiring ~2.3 hours of human labor each, and finds that frontier LLMs are cheaper and faster but still below human quality.

The article argues that AI agents will replace human workers not because they are geniuses, but because much of knowledge work consists of repetitive, menial tasks that are already 'agent-shaped'. It questions what parts of jobs truly require human input.

OfficeCLI is an open-source CLI tool that enables AI agents to read and edit Microsoft Office files (Word, Excel, PowerPoint) without requiring Office installation. It provides a single binary that agents can use to create, modify, and render documents, with built-in HTML rendering for AI vision.

Flexion Robotics, a Swiss startup founded by ex-Nvidia researchers, has developed an AI system that trains humanoid robots to perform complex office tasks by combining simulation, reinforcement learning, and video observation, enabling autonomous operation like retrieving parcels and using elevators.

A detailed guide on how to apply Codex to office and knowledge scenarios, covering environment setup, core concepts, five advanced levels, and other practical content.

The author describes using OpenAI's Codex model to generate actual Office files (.pptx, .docx, .xlsx) directly via function calls, resulting in a useful end-to-end document generation pattern for AI agents.

This paper introduces OfficeEval, a benchmark based on China's National Computer Rank Examination (NCRE) to evaluate LLM agents on complex Office automation tasks. Frontier models achieve at best 36.6% in single-turn and 68.8% with agentic systems, far below human-level performance.

OfficeDex is the first AI-native VibeOfficing platform that generates native .docx/.pptx/.xlsx files directly from natural language, using multiple agents for structure, content, and formatting, with memory and built-in preview.

@dotey suggests Kimi Code and DeepSeek Harness develop GUI and support office tasks early, becoming general-purpose agents, and believes that only competing on TUI and Coding has no future.

Domestic office platforms like Feishu have released open-source CLI toolkits for agent invocation. Feishu has the most features and the highest Star count, surpassing 10,000 Stars in just over 40 days.

OfficeCLI is an open-source command-line tool that lets you create, read, and modify Word, Excel, and PPT files in the terminal without installing Office. It integrates with AI coding assistants like Claude Code and Cursor, making it ideal for automation scripts and batch file processing.