DEV Community: DHIRAJ BHAVIKATTI

Sloth-Sorter: Command Line Tool.

DHIRAJ BHAVIKATTI — Sun, 25 Oct 2026 07:19:01 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady

What I Built

Sloth Sorter is a command-line tool. It sorts and renames the files in your Downloads folder. A small AI model on your own computer does the work. No file leaves your machine.

Example: bill_oct.pdf becomes Documents/Bills/2026-10_Electricity-Bill.pdf.

Many AI organizers upload your bank statements, IDs, and resumes to a server you do not control. Sloth does not. Anyone with a messy Downloads folder can set it up once and stop sorting by hand.

Demo

$ sloth plan --folder ~/Downloads
Plan: 14 files
  bill_oct.pdf  ->  Documents/Bills/2026-10_Electricity-Bill.pdf
  setup.exe     ->  Installers/setup.exe
2 files need review. Nothing has changed yet.

$ sloth run --folder ~/Downloads --yes
14 files tidied. 2 need review.

$ sloth undo --last 3
Put 3 files back.

Code

github.com/DhirajB05/sloth-sorter

How I Built It

Open-source parts

Ollama runs the model on my computer.
Qwen2.5 3B is the default model. It runs on CPU. You can swap it for Llama 3.2 3B or another model with one line in sloth.toml.
Python tools: watchdog, typer, pydantic, pypdf, python-docx, SQLite.

How it works

Watch the folder for new files.
Wait until the file is fully downloaded (size and time stay the same for 3 seconds).
Read the first 1,500 characters of the file.
Ask the model for a category and a clean name. Ollama must reply in a fixed JSON format. Pydantic checks the reply.
Make sure the new path stays inside the folder.
Move the file. Save the move in a SQLite log.

Fallbacks

If Ollama is off, Sloth sorts by file extension. If the model is not sure, the file goes to Needs Review.

Safety rules

Sloth never deletes or overwrites a file. A name clash gets a number: file (2).pdf.
Sloth ignores partial downloads (.crdownload, .part, .tmp).
Sloth only talks to localhost. A test fails if any other network call happens.
Every move can be undone with sloth undo.

Limits

The model makes mistakes. Example: [one real wrong result].
Sloth does not read images. It uses only the file name.
Scanned PDFs have no text, so the model has little to read.

Why Does Open Innovation Matter?

Privacy: your files stay on your disk. A closed API needs an upload for each file.
Offline: it works with no internet. [I tested with Wi-Fi off: result.]
Cost: it costs nothing to run after install. A closed API charges for each call.
Swap models: I compared two models by changing one config line. A closed API gives you the model the vendor chooses.
Control: I set my own folders and naming rules in a config file.

TrailEar: Offline Birding Companion for Hacktoberfest Touch Grass Challenge

DHIRAJ BHAVIKATTI — Sun, 11 Oct 2026 18:34:30 +0000

This is a submission for the** Hacktoberfest** Open-Source AI Challenge Week 1: Touch Grass

What I Built:
TrailEar is an offline birding companion that keeps your eyes on the trail. Most bird-ID apps need signal, hold your attention on a screen, and send your audio and location to a server. TrailEar listens through the mic, gives a quiet cue when it detects a bird, and saves the details for a summary after the walk.

It’s for beginner and casual birders who want to learn calls without staring at a phone.

Demo:
Run locally:
uvicorn trailear.backend.api:app --host 127.0.0.1 --port 8000

Code:
https://github.com/DhirajB05/HacktoberFest_C2.git

FastAPI backend, SQLite storage, and a static HTML/JS frontend with six views (Home, Setup, Live, Summary, History, About).

How I Built It
Detection: BirdNET-Analyzer on 3-second mic windows (sounddevice, 48 kHz mono), with a confidence threshold and a 60-second cooldown per species so cues stay rare.
Summaries: a small local LLM through Ollama or llama.cpp. It only summarizes the detection list plus a static species file, and is told not to add facts.
Live cues: server-sent events, so there is no extra infrastructure.
Privacy by design: the server binds to localhost, there is no telemetry, and there are no runtime network calls.

Why Does Open Innovation Matter?
Works with no signal. A cloud API fails exactly where trails are.
Your data stays yours. Audio and GPS history never leave the device.
Costs nothing to run. There are no per-call fees.
Swappable parts. The LLM or the detection model can be replaced without touching the rest.
Honest limits. Small local models are weaker than closed ones, which is why the LLM only does constrained summaries.

My Agent Session

Built with the opencode agent framework: backend scaffolding, audio capture, the frontend, and end-to-end testing.
Session link:https://opncd.ai/share/STkDsk70

Hi there!, i'm new to DEV community....

DHIRAJ BHAVIKATTI — Mon, 05 Oct 2026 10:00:52 +0000