AirJelly

AirJelly is an always-on desktop AI agent that presents itself as your proactive, self-organizing second brain. Rather than waiting for a prompt, it observes authorized on-screen activity to understand who you are and what you are working on, then proactively captures your intent, organizes your tasks and gets things done. The website describes it as the world's first context-aware proactive agent, built for anyone whose work moves faster than their notes. That includes product, operations and engineering teams, founders and executives, researchers and PhD students, and anyone juggling more context than one brain can hold. AirJelly brings scattered context together so that tools like Codex and Claude Code can act with the full picture, and it also tracks your personal interactions so you never miss a follow-up or forget a promise. The problem AirJelly addresses is the gap between the fast-moving reality of modern work and the static tools people use to track it. Most AI assistants wait for you to ask, and they only understand what you paste into them. Meanwhile, the most important context of a working day — a chat message confirming a review, a note saved weeks ago, a task agreed inside a meeting — stays scattered across tabs, documents, calls and inboxes. Notes apps and to-do lists require you to stop and write things down, which means the fastest-moving details are usually the ones that get lost. AirJelly's premise is that this scattered context is exactly where work slips through the cracks, so the product sets out to observe, remember and act continuously instead of reactively, aiming to free your time for the things AI cannot do — or simply for you to enjoy. The first of AirJelly's three core steps is watching. According to the website, AirJelly captures authorized screen context across your work apps. It reads across every app you use, including Slack, Zoom, Google Docs, Calendar and email, and the FAQ summarises this simply: if it is on your screen, AirJelly can see it. This ongoing observation is what allows the agent to know who you are and what you are working on at any given moment, without you having to explain it. Because the watching happens continuously in the background rather than only when you open an app, the agent builds up a picture of your work as it actually happens. The second step is remembering. AirJelly builds a living timeline of your work in which every task, meeting and message is searchable, forever. In the feature list this appears as the Timeline — a full log of everything AirJelly observed, across every app — and as Memory, which lets you ask about anything from your past work, anytime. The product also describes an activity log and the idea of gathering context for the agent, so that everything it has seen can be brought back when it is relevant. This matters because recall is usually the hardest part of knowledge work: the note you saved, the decision that was made, or the person who promised something are all retrievable rather than lost in a scrollback you will never revisit. The third step is acting. AirJelly surfaces what matters, creates tasks, prepares briefs and sends reminders — before you ask. Tasks are auto-captured from conversations, meetings and messages, complete with due dates, so you do not have to write anything down yourself. Proactive alerts deliver follow-ups, reminders and briefs at the right moment, tying the promise you made in a chat to the moment it needs to be honoured. A Daily Report provides an end-of-day summary of what happened, what is pending and what is next, giving a structured close to the day. Taken together, these features turn observation into output: not just a record of activity, but a stream of concrete items that need attention. AirJelly's overall approach can be described as moving from context to memory, and then from memory to action. The website frames this with three pillars — context for the agent, an activity log, and deadline alerts — and with a three-step loop of watches, remembers, acts. Scattered context is pulled together from the apps you already use, stored as a searchable timeline, and then handed to an agent that can reason over the full picture instead of a partial prompt. AirJelly also explicitly positions itself as a companion for other AI agents, bringing scattered context together so that Codex and Claude Code can act with the full picture. In contrast to regular AI assistants that wait to be asked, AirJelly watches your workflow continuously and acts before you ask, which the site presents as its central differentiator. The benefits AirJelly claims follow directly from that design. You no longer need to take notes to keep track of your own work, because tasks are extracted automatically from conversations, meetings and messages with their context and due dates attached. You are less likely to drop a follow-up or forget a promise, because personal interactions are tracked and reminders arrive at the right moment. You spend less time re-reading documents or hunting through open tabs, because a one-page summary and a searchable memory can surface what you need — including notes you had forgotten you saved. And when an important moment is coming up, such as an interview, a brief is already prepared. The stated outcome is that everything in your life is managed by AirJelly, freeing your time for what AI cannot do, or simply for you to enjoy. The website illustrates this with concrete scenarios. In a product and operations setting, a team confirmed a review in a chat message; AirJelly caught it, created the task and attached the conversation, with the due time set and the context attached. In research and academia, a user had fourteen tabs open across papers, docs and notes; AirJelly read through all of it and produced a single research summary, including a note the user had forgotten they had saved, with key points extracted. For founders and executives, an interview thirty minutes away was spotted on the calendar and AirJelly prepared a brief covering background, recent work and talking points. These examples all follow the same pattern: context is observed, kept, and then converted into something useful before the user asks for it. AirJelly is available as a free download for Apple Silicon Macs and Windows x64 PCs, with Linux support stated as coming soon, and the site invites users to download and start in minutes, as well as to join its Discord community. It works with the apps already on your screen — Slack, Zoom, Google Docs, Calendar, email and more — rather than requiring a new place to work. On privacy, the site states that screenshots are encrypted and stored on your device, that capture can be paused at any time, and that specific apps can be excluded, with further privacy controls documented on a dedicated page. In summary, AirJelly's primary value proposition is proactivity: an always-on desktop agent that watches your workflow, remembers everything in a searchable living timeline, and acts before you ask by creating tasks, preparing briefs and sending reminders. For teams, founders, researchers and anyone carrying more context than one brain can hold, it aims to close the gap between the work that is happening on screen and the notes that never quite get written.