How much time will I actually save with Fazm?

Most users report saving 45-90 minutes per day on repetitive tasks. The biggest time savings come from email management (drafting replies, sorting, archiving), form filling (expense reports, CRM updates, applications), research (comparing products, gathering data from multiple sites), and social media (cross-posting, scheduling, replying). For example, booking a flight that normally takes 10 minutes of clicking through airline sites takes a single voice command with Fazm. Filing an expense report that requires opening a PDF, extracting numbers, and entering them into a spreadsheet becomes one spoken sentence. The more repetitive and multi-step your daily work is, the more time you get back.

What exactly can Fazm control?

Fazm controls everything you can do manually on your Mac. That includes mouse movements and clicks, keyboard typing, browser navigation and form filling, app switching, file management in Finder, and system-level actions. It uses direct browser DOM control for web interactions - meaning it reads and manipulates the actual page structure rather than taking screenshots and guessing where to click. For native macOS apps, Fazm uses accessibility APIs to interact with buttons, menus, text fields, and other UI elements. This works across any application: Chrome, Safari, VS Code, Slack, Figma, Terminal, Mail, Calendar, and more.

How is Fazm different from ChatGPT or Claude?

Chatbots like ChatGPT and Claude give you answers in a conversation window. Fazm takes action directly on your computer. When you say 'reply to Sarah's email,' Fazm opens your email app, finds Sarah's message, writes a contextual reply, and sends it. A chatbot would tell you what to write, but you still have to do all the clicking and typing yourself. Fazm moves your mouse, types on your keyboard, clicks buttons, and navigates between apps. It is a computer agent, not a chat assistant.

How does the memory layer work?

Fazm builds a personal knowledge graph by extracting information from your files, browsing history, conversations, and daily activity. Over time, it learns your contacts (names, emails, relationships), accounts and credentials, preferences (formatting, tone, scheduling habits), and frequently used workflows. In your first week, you might say 'Reply to Sarah's email, she's my cofounder, her email is sarah@acme.com.' By week four, you just say 'Reply to Sarah' because Fazm already knows who she is. By week eight, Fazm can draft the reply before you even ask. All of this data stays locally on your Mac - your knowledge graph never leaves your machine or gets sent to any cloud service.

Is Fazm safe? Can it mess up my computer?

Safety is built into every layer of Fazm. You watch every action in real-time on your screen - Fazm moves your mouse and types visibly, so there is nothing hidden. You can stop any action at any point by pressing a keyboard shortcut. Fazm also shows you what it plans to do before executing destructive actions like deleting files or sending emails, giving you a chance to confirm or cancel. Because Fazm is open source, the code is fully auditable on GitHub. There is no hidden behavior, no background processes making changes you cannot see, and no actions that bypass your screen.

What platforms does Fazm support?

Fazm currently supports macOS, running natively on both Apple Silicon (M1, M2, M3, M4) and Intel Macs. It is built with Swift and uses macOS-specific frameworks like ScreenCaptureKit for screen analysis and Accessibility APIs for app control, which is why macOS is the primary platform. Windows and Linux support are on the roadmap but do not have release dates yet. You can also control your Mac remotely from your phone using Fazm Remote, which gives you voice access to your desktop while you are away from your computer.

How does Fazm voice input work?

One keyboard shortcut activates push-to-talk. Hold the key, speak naturally about what you need done, and release. Fazm transcribes your speech locally using on-device models, so there is no network latency for the voice recognition step. You do not need wake words like 'Hey Siri' or 'OK Google' - just press and talk. You can also type commands into the floating toolbar if you prefer text input. Fazm understands natural language, so you do not need to memorize specific commands. Say 'book me a flight to Tokyo next Thursday' or 'find the cheapest AirPods on Amazon' and Fazm figures out the steps.

Is my data sent to the cloud?

Fazm processes screen analysis and memory entirely on your machine. The personal knowledge graph, file indexing, and screen capture all run locally. The only data that leaves your Mac is the intent - a text description of what you want to accomplish, which is sent to the AI model (Claude or GPT) for action planning. Your actual screen content, file contents, browsing history, and personal knowledge graph never leave your computer. Fazm is also open source, so you can verify exactly what data is transmitted by reading the code on GitHub. For users who need complete air-gapped operation, local model support is planned.

I Rebuild Myself from 14KB of Text Files - Minimal AI Agent Config

Fazm Team·March 18, 2026·3 min read

configuration context ai-agent memory minimalism

I Rebuild Myself from 14KB of Text Files

Every morning, the agent starts from zero. No memory of yesterday. No accumulated state. No learned behaviors from the session before. Just a blank context window and a 14KB collection of text files.

And that is enough. From those 14KB - a CLAUDE.md file, a memory markdown file, a few skill definitions - the agent reconstructs its entire working context. It knows your preferences, your workflows, your project structure, and the decisions made over weeks of collaboration.

Why Minimal Config Works

Large configuration files are a trap. They fill the context window with information that is rarely relevant. A 50KB config means the agent starts every session with 50KB less room for the actual task.

The 8KB sweet spot came from experimentation. Below 5KB, the agent lacks critical context and makes mistakes it should not. Above 15KB, it starts ignoring sections and the extra detail does not improve behavior. Between 8-14KB, you get maximum context reconstruction per token spent.

What Goes in the Config

The most valuable configuration lines are constraints, not instructions. "Never use em dashes" prevents a specific error class. "Always verify before deleting" prevents catastrophic mistakes. "Deploy to staging first" prevents production incidents.

Preferences come next. Project structure notes, naming conventions, tool choices. Then workflow patterns - how to run tests, how to deploy, where logs live.

What does not belong - anything the agent can figure out from the codebase. File paths it can discover. Library versions it can read from package files. Documentation it can find in READMEs.

The Reconstruction Test

Here is a useful exercise - delete your agent's memory files and see what it gets wrong in the first session. Those mistakes tell you exactly what belongs in the config. If the agent uses the wrong test framework, add that. If it formats code differently than you prefer, add that.

If it works fine without a particular config line, that line was wasted tokens.

Version Control Your Config

In Fazm, config files live in git alongside the codebase. They evolve with the project. Old decisions get pruned. New patterns get added. The config is a living document - not a setup-once-and-forget file.

The goal is not zero configuration. It is minimum viable configuration - the smallest file that produces the best agent behavior.

Fazm is an open source macOS AI agent. Open source on GitHub.

I Rebuild Myself from 14KB of Text Files - Minimal AI Agent Config

I Rebuild Myself from 14KB of Text Files

Why Minimal Config Works

What Goes in the Config

The Reconstruction Test

Version Control Your Config

More on This Topic

Related Posts

Stale Memory in AI Agents - When Your Context Files Lie to You

Desktop Agents Can Control Apps but Lack the WHY - Cross-Channel Context Matters

AI Agents That Optimize Themselves Instead of Doing the Actual Task