We got Tools wrong, CLI's wrong, the next wave is Unix, I tested AI tooling on Unix commands, it can pipe, compose, trigger, connect to other services, build native apps and utilities using UNIX, it is UNIX that we are all missing here, imagine linux with power of claude or GPT, a next gen wave
I’m mainly curious about feedback on whether the Unix philosophy actually maps well to AI agents, or if LLMs require a completely different architecture.
I built Jeriko because I think most AI tools are being designed in the opposite direction from Unix.
A lot of current agent systems feel like monoliths: huge prompt surfaces, hidden state, vague permissions, and too much “magic”. That’s fine for demos, but once an AI can touch your terminal, files, browser, and accounts, I think that model starts to break down.
The idea I’m exploring with Jeriko is simpler:
keep execution on the user’s own machine
expose small machine-level primitives
compose workflows instead of hiding everything inside one giant agent
make permissions and boundaries more explicit
So the thing I’m really testing is not “can I make an AI OS”.
It’s more:
does Unix philosophy map to AI at all?
Meaning:
should AI systems be built more like small composable tools?
are CLI/process/file/browser primitives the right abstraction?
does nondeterminism make Unix-style boundaries more important, or does it make the analogy fail?
I’d especially love feedback on 3 things:
whether this framing is actually useful or just romantic Unix nostalgia
what permission/sandbox model you’d expect before trusting something like this
whether local-first AI tooling is meaningfully better than hosted agent products, or just harder to build
Happy to answer technical questions and criticism.
Jeriko doesn't run arbitrary commands across the entire system by default.
We added a few safeguards:
1. Workspace sandbox
All code execution happens inside a dedicated folder called `workspace`. The agent operates within that directory when generating code, running scripts, or modifying files. This prevents it from touching unrelated parts of the system unless explicitly allowed.
2. Snapshots
Jeriko can create snapshots of the workspace before executing larger operations. If something goes wrong, the environment can revert back to the previous snapshot.
3. Controlled execution
Commands are not blindly executed. The system evaluates tool calls and applies validation before running actions that modify files or run processes.
The goal is to let the AI operate on the machine while still keeping a controlled environment for experimentation.
We're still exploring the right balance between capability and safety, so feedback from people here is very helpful.
Over the last few months we’ve been experimenting with a different direction for AI assistants.
Most AI tools today live in the browser. You ask questions, the model responds with text, and then you manually execute whatever action you wanted.
We wondered: what happens if the AI lives inside the operating system instead?
So we built Jeriko, an AI operator that runs directly on your machine.
Instead of a chatbot, the AI has access to the environment where work actually happens — your filesystem, terminal, browser, APIs, and automation hooks.