InfinityAI - V2 banner

InfinityAI - V2

Created by: srinivasagudi0
6 devlogs
73h 6m 34s

Updated Project: After shipping the project last time, I realized the project didn’t clearly show what it could actually do and parts of it didn’t feel fully like my own work. So in this version, I focused on fixing that and making everything more…

Updated Project: After shipping the project last time, I realized the project didn’t clearly show what it could actually do and parts of it didn’t feel fully like my own work. So in this version, I focused on fixing that and making everything more real and understandable.
I added a live system log so you can see what the AI is doing step by step, and changed the system so it executes tasks instead of just explaining them. I built watch mode for both CLI and UI (with a separate UI screen and a better CLI using rich), added Supabase authentication for real users, and fixed a lot of core issues like state handling, device handling, errors, timeouts, and overall stability. I also reworked parts of the codebase to make it more structured and truly my own.

This project uses AI

I used codex to write README and help me with the UI.

Demo Repository

README

Loading README...
srinivasagudi0
srinivasagudi0 worked on InfinityAI - V2
about 4 hours ago
32h 44m logged

I updated and refined the user interface to present a design that is both simpler and more futuristic, while retaining all the original functionalities. Additionally, I will address and resolve any errors that may persist to ensure optimal performance.

Attachment
0
srinivasagudi0
srinivasagudi0 worked on InfinityAI - V2
about 7 hours ago
0h 15m logged

I have updated and cleaned the GitHub repositories, and I am continuing to work on the Streamlit user interface, which has proven to be quite challenging.

Attachment
0
srinivasagudi0
srinivasagudi0 worked on InfinityAI - V2
about 8 hours ago
4h 55m logged

Added file integration to the UI and next I will try to make it better.

Attachment
0
srinivasagudi0
srinivasagudi0 worked on InfinityAI - V2
1 day ago
22h 59m logged

What’s accomplished:

typing file in the CLI opens a dedicated file mode
it asks for file paths and loads them
it shows a polished Rich screen with:
header
loaded files list
session/chat area
command help
you can ask normal questions about the loaded files
it supports:
/list
/add
/remove
/reload
/summary
/fix
/help
/exit
Also:

missing files do not crash the workspace
unreadable files do not crash the workspace
unsupported file types are shown clearly
PDF/image support no longer breaks CLI import if optional libs are missing

Attachment
0
srinivasagudi0
srinivasagudi0 worked on InfinityAI - V2
1 day ago
8h 28m logged

I set up Supabase-based user authentication for login and account creation, gave voice mode a major upgrade with improved audio device handling, better session states like “transcribing,” clearer prompts, stronger error handling, and reply timeouts, and also refreshed the Infinity Command Core UI while updating the README and shipping/version docs.

Attachment
0
srinivasagudi0
srinivasagudi0 worked on InfinityAI - V2
1 day ago
3h 42m logged

I reviewed the issues from my last project submission and made improvements. Now, when Infinity runs, it displays a visible thinking process instead of just a generic spinner. In short, the CLI, Voice UI, and Web UI all show an Agent Thinking Log with steps like understanding the prompt, planning files, generating code, and fixing errors, making it feel like a real system working step by step rather than just a GPT wrapper saying “thinking.” It’s now smarter based on the task—showing planning and code generation for builds, search-related steps for searches, and reply-related steps for normal chat. I also fixed a backend issue so that if the project planner fails, Infinity keeps going instead of stopping. All related tests have passed.

Attachment
0