Same clicks.Same tabs.On repeat.
The little tasks have a way of taking up your whole day. Give them a different kind of shortcut.
Ask your computer. Give it context. Put it to work.
Meet Jarvis, your local-first AI assistant. Source included.
The little tasks have a way of taking up your whole day. Give them a different kind of shortcut.
Search, files, browser, code. Bring the things you do into one conversation.
Your preferences deserve more than a copy and paste. Give your assistant a memory.
Your next shortcut is a conversation.
“Let’s simplify the navigation before launch.”
from pathlib import Path for file in Path("Downloads").iterdir(): # preview before making changes print(file.suffix, file.name)
A little context for your next conversation.
“Find the project brief in this folder.”
Right where you left it.“Transcribe this recording. Pull out the key points.”
A thought, made useful.“Help me write a script to organize these files.”
From idea to a first draft.“Remember that I prefer concise answers.”
Less explaining. More you.“Let’s simplify the navigation before launch.”
from pathlib import Path for file in Path("Downloads").iterdir(): # preview before making changes print(file.suffix, file.name)
A little context for your next conversation.
Voice or text. Start however you like.
Memory makes the next request personal.
File, browser, and terminal tools.
Scroll through. Play a little.
Find your kind of useful.
Interactive illustrations. Explore each capability for setup and availability.
45 capabilities. 30 practical walkthroughs. Follow an ordinary request into something surprisingly useful.
A project to revisit. A thought to come back to. Give it a time, and let Jarvis keep the schedule.
Prepare a project brief every weekday. Turn a one-time reminder into a recurring agent task, in the same conversation.
Pause, resume, or remove a scheduled task. Your local Jarvis scheduler needs to stay running to carry it out.
A few things your next conversation could become.
“Jarvis, transcribe this recording and summarize the key points.”
Explore this possibility“Jarvis, search the web and walk me through what you find.”
Explore this possibility“Jarvis, help me write a script to organize these files.”
Explore this possibility“Jarvis, remember that I prefer concise answers.”
Explore this possibility“What if we made a little more room for the things we want to make?”
“Jarvis, transcribe this recording and summarize the key points.”
Explore this possibility“Jarvis, search the web and walk me through what you find.”
Explore this possibility01 from pathlib import Path 02 03 for file in Path("Downloads").iterdir(): 04 print(file.suffix, file.name)
.md project-brief.md
.py idea.py
“Jarvis, help me write a script to organize these files.”
Explore this possibilityShort. Clear. Just enough.
“Jarvis, remember that I prefer concise answers.”
Explore this possibilityOne assistant.
Wherever the idea goes.
Connect a cloud provider or an installed local model server. The provider, model, and connection are yours to configure.
Connect configured MCP servers to discover and call their tools. Each service keeps its own setup and authentication.
Add reusable skills, Python extensions, and specialist agents. Local models need installation; cloud services and online tools need a connection.
Here's a little more context.
Get the source. Bring your providers. Turn Jarvis into the assistant that works the way you do.