lsv is a fast curses-based terminal file viewer with 3 side-by-side panels for parent, current and preview.
Parent shows the parent directory, Current handles selection and navigation in the active directory, and Preview shows the selected file or directory contents.
It’s keyboard-driven, configurable with Lua, and supports ANSI-colored previews from tools like bat and glow.
@githubforks
Parent shows the parent directory, Current handles selection and navigation in the active directory, and Preview shows the selected file or directory contents.
It’s keyboard-driven, configurable with Lua, and supports ANSI-colored previews from tools like bat and glow.
@githubforks
Constrict compresses video to a chosen file size for uploads to services with size limits.
The app automates the usual trial and error of re-encoding with different bitrates. It calculates average bitrate, resolution, frame rate and audio quality for each video to hit a target size. It also supports batch compression of multiple videos into one output folder.
Constrict offers frame-rate limits to trade off smoother motion against sharper images. Users can also pick output codecs, including H.264, HEVC, AV1 and VP9. The project is available on GitHub.
@githubforks
The app automates the usual trial and error of re-encoding with different bitrates. It calculates average bitrate, resolution, frame rate and audio quality for each video to hit a target size. It also supports batch compression of multiple videos into one output folder.
Constrict offers frame-rate limits to trade off smoother motion against sharper images. Users can also pick output codecs, including H.264, HEVC, AV1 and VP9. The project is available on GitHub.
@githubforks
OpenSpec is a GitHub project that brings spec-driven development (SDD) to AI coding assistants.
@githubforks
@githubforks
Google is launching a free AI agents intensive starting on June 15.
The program covers how chatbots evolve into full agents with planning and tools, how agents interact with each other and users, context, memory and history storage, plus safety, limits, debugging and production rollout. You can get a certificate after completion. Course
@githubforks
The program covers how chatbots evolve into full agents with planning and tools, how agents interact with each other and users, context, memory and history storage, plus safety, limits, debugging and production rollout. You can get a certificate after completion. Course
@githubforks
Flint is a standalone KVM management tool for developers, system administrators and home lab users who want efficiency without extra features.
It was built in a few hours.
@githubforks
It was built in a few hours.
@githubforks
This media is not supported in your browser
VIEW IN TELEGRAM
Reddit is hyping a site that works like a “TikTok for smart people”: an endless feed of puzzles, with one task replacing the next without stopping.
It cycles through Wordle, chess, Tetris, solitaire, sudoku and dozens more formats. More than 20 genres are mixed at random on the site.
@githubforks
It cycles through Wordle, chess, Tetris, solitaire, sudoku and dozens more formats. More than 20 genres are mixed at random on the site.
@githubforks
Codebases, meeting notes, inboxes, Slack, PDFs and video can be turned into continuously updated context with CocoIndex.
It is pitched as a way for AI agents and LLM apps to make decisions with minimal extra processing, using data that stays fresh across many sources instead of relying on stale batch syncs.
@githubforks
It is pitched as a way for AI agents and LLM apps to make decisions with minimal extra processing, using data that stays fresh across many sources instead of relying on stale batch syncs.
@githubforks
VideoPipe is a C++ framework for video analysis and structuring built as a modular pipeline: each node is independent and can be combined in different ways.
It has minimal dependencies and is described as easy to use for custom video pipelines such as video structuring, image search, face recognition, and traffic and safety behavior analysis, including road accident detection.
@githubforks
It has minimal dependencies and is described as easy to use for custom video pipelines such as video structuring, image search, face recognition, and traffic and safety behavior analysis, including road accident detection.
@githubforks
Reconya turns local network discovery into live monitoring by detecting and tracking devices in real time.
Written in Go, Reconya is a network reconnaissance and asset discovery tool with real-time updates for admins, security teams and regular users.
@githubforks
Written in Go, Reconya is a network reconnaissance and asset discovery tool with real-time updates for admins, security teams and regular users.
@githubforks
Not limited to one modern format, XL Converter supports JPEG XL, AVIF, JPEGLI, lossless JPEG transcoding, fast parallel encoding, and ExifTool.
@githubforks
@githubforks
NoctaVox is a lightweight Rust music player with a TUI for local libraries and terminal use: NoctaVox.
@githubforks
@githubforks
BentoPDF edits, merges and processes PDF files in the browser, with no uploads and no server-side processing.
BentoPDF is a privacy-focused client-side PDF toolkit. It does not require hosting and handles PDF manipulation, editing, merging and processing locally, which keeps files on the device.
@githubforks
BentoPDF is a privacy-focused client-side PDF toolkit. It does not require hosting and handles PDF manipulation, editing, merging and processing locally, which keeps files on the device.
@githubforks
Agentic-rag-for-dummies shows how to build a modular agentic RAG system with LangGraph using minimal code.
Most RAG guides cover the basics, but not how to structure modular agent systems. This project fills that gap with learning materials and an extensible architecture.
@githubforks
Most RAG guides cover the basics, but not how to structure modular agent systems. This project fills that gap with learning materials and an extensible architecture.
@githubforks
OSINT Framework helps people find free OSINT sources for gathering information with free tools and resources.
Some listed sites may require registration or offer more data for $$$, but users should still be able to access at least part of the available information for free.
@githubforks
Some listed sites may require registration or offer more data for $$$, but users should still be able to access at least part of the available information for free.
@githubforks
Serie is a TUI app that uses terminal image display protocols to render commit graphs such as
@githubforks
git log --graph --all.@githubforks
Docus is a Nuxt CLI that scaffolds a full Markdown documentation site with a clean docs theme built on Nuxt UI 4 and Tailwind CSS 4.
You get a mobile-first responsive layout, built-in light and dark mode, 20+ locales for the support UI, client-side search with an optional FTS5 full-text backend, MDC custom Markdown components, theme options, custom icons, Nuxt 4 performance work and full TypeScript support.
@githubforks
You get a mobile-first responsive layout, built-in light and dark mode, 20+ locales for the support UI, client-side search with an optional FTS5 full-text backend, MDC custom Markdown components, theme options, custom icons, Nuxt 4 performance work and full TypeScript support.
@githubforks
Want to run LLMs locally instead of relying on cloud APIs? Awesome local LLM collects platforms, tools, practices and resources in one place.
@githubforks
@githubforks
Ohmycaptcha is a self-hosted CAPTCHA solver compatible with YesCaptcha, built with FastAPI, Playwright, and OpenAI-compatible multimodal models.
Ohmycaptcha
@githubforks
Ohmycaptcha
@githubforks
SvgPathEditor is a browser-based editor for creating and managing SVG paths.
SvgPathEditor
@githubforks
SvgPathEditor
@githubforks
PDF Document Layout Analysis reportedly analyzes PDF layouts as structured parts rather than flat pages.
The Docker-based service segments and classifies page elements including text, headings, images, tables and more.
@githubforks
The Docker-based service segments and classifies page elements including text, headings, images, tables and more.
@githubforks
Sokuji does live two-way voice translation for bilingual meetings with automatic language detection, translating in both directions.
It runs in the cloud or fully offline on-device, with desktop apps for Windows, macOS and Linux, plus a Chrome and Edge extension for Zoom, Meet, Teams and other apps.
@githubforks
It runs in the cloud or fully offline on-device, with desktop apps for Windows, macOS and Linux, plus a Chrome and Edge extension for Zoom, Meet, Teams and other apps.
@githubforks