tierlist.tech
中EN

09.10 Today's Picks

406 new projects found today, 219 rated.

Created 2026-09-08 – 2026-09-10Fetched 600Filtered 50Rated 30

S · Look today

S

glooprint

ishtmsC++

An Unreal Engine plugin that auto-lays out and rewires Blueprint node graphs with one keypress, tidying the whole graph. Messy Blueprints are a daily pain for UE developers, so the angle is small and precise; limits are it only serves Blueprints, and stability needs testing in the editor.

An Unreal Engine plugin for automatic Blueprint graph formatting and clean/performant wire routing. Press F to tidy up your entire graph.

#Game Dev#Plugins#Editors#unreal-engine

★26

A · Worth watching

A

sonascope

ranshufang55TypeScript

A browser audio visualization workbench that is also a TypeScript DSP library: waveforms, spectra, spectrograms, WAV conversion and streaming analysis. People debugging web audio or building frontend audio features can run it the same day. The limit is it still trails Audacity and Sonic Visualiser by a wide margin, and documentation maturity matters.

A local audio visualization workbench and TypeScript DSP library with waveform, spectrum, spectrogram, Web Audio, WAV interchange and streaming analysis.

#Audio#SDKs#Web Frontend

★31

A

nervous-system-atlas

aycibatuhanPython

A browser-based 3D clinical neuroanatomy atlas: 592 meshes, synchronized MRI slices, vascular territories and tracts, case modes and quizzes, with entries citing open literature, in English and Turkish. A genuinely usable tool for medical students and educators; the downside is heavy frontend and data with high maintenance cost.

A browser-based 3D atlas of clinical neuroanatomy in MNI152 space: 592 meshes with synchronised MRI slices, arterial territories, traced pathways, a syndrome and lesion mode, clinical topics, a glossa…

#Science#Biotech#3D#Health#Learning

★21

A

dst-character-studio

0xleonvibePython

A character art workbench for Don't Starve mod authors: layered drawing, rig calibration, animation preview and export to build, complete around one task. The positioning and description are honest, without jargon; limited to the small DST circle, only a day old, and actual completeness needs installing to judge.

A local art workbench for Don't Starve Together characters: layered drawing, binding calibration, native animation review and build export.

#Games#Game Dev#Animation#Editors

★14

A

effect-clickhouse

MapleTechLabsTypeScript

A ClickHouse client for the Effect ecosystem: type-safe queries, result decoding, plus reproducible benchmarks. The scope is restrained and the naming honest, filling Effect's gap in OLAP scenarios; the limitation is that it only matters to teams already using Effect.

Type-safe ClickHouse queries, result decoding, and reproducible benchmarks for Effect and TypeScript

#Databases#Performance#SDKs#clickhouse

★18

A

appeditions

mirairoadGo

Batch-generates localized app store screenshots locally and tracks per-version, per-language progress; no account, no uploads. Such assets are mostly paid online services, so a local Go desktop app is a clear difference. The limit is it only covers the narrow iOS/Android store asset case, and template and review-size fit still need verification.

Turn raw app screenshots into store-ready App Store and Google Play assets, and keep tracking of the version you are in, in every language you ship in. A local Go desktop app — no account, no upload, …

#Images#Productivity#app-store

★15

A

self-hosted-ci-runner

senoffShell

Starts multiple GitHub Actions runners on one machine with a single command, directly addressing heavy CI filling disks and OOM: each runner gets its own swap and scratch plus cleanup scripts, with a Hetzner reference config. Teams running self-hosted CI can use it directly; the limit is it is more an ops script than a general solution.

Stand up self-hosted GitHub Actions runners on one box, provisioned so heavy CI jobs don't OOM or fill the disk (per-runner swap/scratch + reaper). Parameterized; Hetzner reference config included.

#CI#DevOps#Self-Hosted#Automation

★18

A

voxweave

chenkb98Python

Extracts the dirtiest glue work around streaming speech LLMs into a separate project: audio frames, turns, session schemas, batching, and vendor adapters. People building real-time voice agents can save a layer of glue; the repo is under a day old with no examples, so interface stability needs watching.

Streaming speech language model plumbing: audio frames, turns, conversation schemas, batching, and adapters.

#Speech#LLMs#Audio#Multimodal

★32

A

sonar.cool

eperez28Swift

Uses a Mac's built-in speakers and microphone as 'sonar' to control scrolling, switching, and zoom via gestures, all without a camera. A rare sensing approach, and the author labels it experimental without overselling; limits are that it only recognizes simple gestures, with limited precision and scenarios.

Control your Mac with hand gestures using its speakers and microphone. Experimental, camera-free scrolling, swiping, and zooming.

#macOS#Audio#Hardware#Accessibility

★13

A

wsl-reclaim

ttncodeShell

WSL2 disks that only grow are an old problem: clean up inside the distro first, then compact the vhdx on the Windows side, and it also explains why diskpart silently fails on sparse disks. Directly relevant to WSL users on Windows. The limitation is that it is a pure shell script relying on manual confirmation, winning only by explaining the principle clearly.

Reclaim WSL2 disk space for real: clean inside the distro, then compact the .vhdx on Windows. Explains why diskpart silently does nothing on sparse disks.

#Windows#Linux#CLI#DevOps

★12

A

mmd-mcp

BeamManPPython

Exposes MikuMikuDance modeling and MME settings as a stdio MCP server on Windows, letting LLMs drive MMD directly — a rare interface. The scope is clearly stated; limited to MMD users on Windows, and how much it actually covers needs testing.

Windows stdio MCP server for MikuMikuDance authoring and MME settings

#MCP#3D#Animation#Windows#mmd

★21

A

polymo

brenpolyKotlin

An electronic pet: ESP32-S3 firmware plus an Android app, with all 'thinking' done locally on the device; the hardware plus on-device inference combo is concrete enough for DIYers to follow. The limit is that it's only a virtual pet toy, and reproducibility and long-term maintenance need more days to judge.

A physical digital pet you keep alive — ESP32-S3 firmware and an Android app that does all the thinking on-device.

#Embedded#Hardware#Android#IoT

★25

That's today. The other tiers and the boards follow.

BHas a spark52
  • B anything2explainer Turns any topic into a generated explainer video with voiceover and subtitles, as a Claude Code/Codex skill. It gained 76 stars in 1.3 days, showing demand for automation in content creation; commercialization means a paid skill or template for vloggers and course makers, but video generation tools change fast and competition is fierce. #AI Agents#Agent Skills#Video Editing#Content Writing

    Topic in, narrated explainer video out. A Claude Code / Codex skill that turns any topic into a black-canvas motion-graphics explainer video with TTS voiceover, subtitles and a chapter progress bar. C…

    Vincentwei1021TypeScript★76

  • B spotkit A Claude Code skill that converts feature descriptions into SVG illustrations for a unified design system. It gained 49 stars in 1.3 days, fitting the agent skill trend; monetization is paid design-skill templates or integration into design tools, with the risk that SVG illustration is limited and users may prefer direct image-generation AI. #AI Agents#Agent Skills#UI Design#Image Generation

    Spotkit — a Claude Code skill that turns a feature description into minimal, abstract SVG product illustrations that all belong to one design system.

    Devesh-ShirsathHTML★49

  • B haiming-app-monetization An AI skill helping developers commercialize apps, such as onboarding, paywall and in-app purchase design. It gained 79 stars in 0.6 days, showing the agent skill space is hot; monetization is a paid skill or integration into agent marketplaces, with the risk of being covered by general AI instructions or lacking systematization and thus limited value. #AI Agents#Agent Skills#Marketing#Productivity

    海明Dev: App monetization Skill | onboarding, paywall, in-app purchase options, competitor research

    HammingDevLanguage unknown★79

  • B INSLIB An inertial navigation algorithm library for embedded fields such as robotics and drones. 188 stars in 1.6 days shows real demand, but inertial navigation already has mature solutions and limited innovation. The commercialization path is as an underlying dependency for hardware and software companies or enterprise support; the risk is that existing commercial software or open-source libraries hold the market, and domain knowledge is required. #Embedded#Science#Hardware

    Open Source Inertial Navigation Library

    jnzC★188

  • B fast-rtxvsr Turns RTX video super-resolution into a standalone command-line single-pass pipeline: GPU decode, VSR, and encode in one go, no ComfyUI needed. RTX users wanting to batch-upscale old videos can run it today; limits are it's locked to Nvidia cards and offers no quality comparisons. #Video#CLI#GPU#nvidia

    Standalone NVIDIA RTX Video Super Resolution CLI - GPU decode -> VSR -> encode in one pass, no ComfyUI

    glarssonPython★11

  • B rubric A TypeScript schema-first structured extraction library that pulls LLM output according to a given schema, with a one-line purpose and restrained interface. It is a lightweight option for developers who need reliable JSON output. Mature alternatives like instructor and outlines already exist, so whether it is worth switching depends on validation and error-handling details. #LLMs#SDKs#APIs

    Schema-first structured extraction from LLMs.

    tomnioTypeScript★26

  • B headroom A closed-loop agent system that judges audio mixes using DSP metrics instead of subjective feel, aimed at audio engineering. It gained 47 stars in 2.0 days; commercialization means becoming a production tool or plugin, but professional audio software already exists and automated mixing struggles to win expert approval. #Audio#AI Agents#Automation

    A closed-loop agent system that mixes and masters audio toward a measured target. Orchestration graded by DSP measurement, not vibes.

    KianSalemPython★47

  • B nibble A Go library for splitting long text into retrieval chunks, clearly scoped and restrained, usable today for Go-side RAG. The knock is that chunking has plenty of existing solutions, from langchaingo to various semantic splitters, and the repo doesn't explain how its splitting strategy differs. #RAG#LLMs#SDKs

    A Go library for splitting long text into retrieval-sized chunks for RAG.

    bluesky585Go★21

  • B omarchy-in-omarchy Spins up a self-installing Omarchy virtual machine with one QEMU/KVM command, skipping the wizard, password, and login, and discarding it after use. Perfect for people who want to try an Arch-based desktop without dirtying their main machine; the limit is it only makes sense for this one distro. #Linux#Automation#Operating Systems#qemu

    A disposable Omarchy machine in QEMU/KVM that installs itself: no wizard, no disk passphrase, no login prompt

    jankeesvwShell★13

  • B Surge A contribution sprint platform for scheduled funding in the Stellar ecosystem, where maintainers post tasks with points and contributors apply with proposals to earn USDC rewards. 93 stars in 1.3 days shows community interest, but the market is limited to Stellar. The commercialization path is becoming the ecosystem's standard incentive mechanism or expanding to other chains; the risk is competition from Gitcoin-like players and the ecosystem's limited size. #Blockchain#Fintech#Collaboration

    Funded, time-boxed contribution sprints for the Stellar ecosystem. Maintainers post scoped issues with a point value, contributors apply with a plan, and accepted work splits the wave pool in USDC. Fr…

    Surge-OrgTypeScript★93

  • B motion-web Builds motion-first creative websites with Three.js/WebGL as a Claude Agent skill, including a physics engine and automatic validation. It gained 43 stars in 1.8 days, meeting frontend creative demand; commercialization means a skill-market product or helping agents generate more motion, but competition is fierce and debugging complex. #AI Agents#Agent Skills#Web Frontend#Animation

    Build motion-first, high-touch creative websites on vanilla Three.js / Canvas 2D / WebGL / CSS. A Claude Agent Skill with real physics and automated verification.

    feitangyuanHTML★43

  • B naming-scout After naming a product, automatically checks domains, trademarks, GitHub, npm, app stores and search, as an agent skill for Claude Code and similar. Closer to the dev workflow than sites like namechk; trademark and store result accuracy still needs testing. #Agent Skills#Productivity#Automation#CLI

    Names your startup, product or side project, then proves you can actually use it. Checks the domain registry, trademark, GitHub, npm, app stores and search. An agent skill for Claude Code, Cursor, Cod…

    me-shaonShell★16

  • B agent-html A zero-dependency single-file HTML design system paired with a skill for LLMs, aiming to have models directly produce pages that open offline. The idea is practical, but the AI HTML template layer is already crowded, and it doesn't explain how its layout prototypes beat the others. #Web Frontend#Agent Skills#UI Design

    Zero-dependency, single-file HTML design system & agent skill inspired by shadcn/ui. 100% offline, zero npm, 6 layout archetypes & atomic arsenal for LLMs.

    QingYunAHTML★22

  • B ai-level-check It reads local agent logs and grades your AI usage from LV0 to LV5, explicitly saying it doesn't rank, with a restrained tone. It's a clean little toy that does one thing and explains it clearly; the limitation is that one look is enough, and you still have to judge the grading criteria yourself. #AI#AI Agents#Productivity

    AI 幾級了 — What level is your AI usage? Reads local agent logs and writes an LV0–LV5 report. No ranking.

    Raymondhou0917Python★26

  • B URBER A reference implementation of a paper using a small-scale rule-based method to solve escape routing for biochip sample transport, narrow in scope, clear in goal, and not oversold. It's directly usable by researchers in EDA and microfluidics; the limitation is that the repo only provides code, so you still have to read the paper to get it running. #Science#Biotech#Hardware

    Code for URBER: Ultrafast Rule-Based Escape Routing Method for Large-Scale Sample Delivery Biochips

    Trinkle23897Python★13

  • B fantasy-football-manager Wraps ESPN fantasy football drafting and lineup setting as an MCP service for clients like Codex, with login state, approval mode and per-team limits. A real need for people who both play fantasy and use agents; the risk is reliance on third-party site APIs that may break at any time. #MCP#AI Agents#Automation#Games

    ESPN draft and lineup tools for Codex and other MCP clients, with browser sign-in, approval modes, per-team limits, and server monitoring.

    krmisystemsPython★32

  • B whiteboard-explainer Turns handwritten whiteboard explanations into frame-aligned animation: shapes drawn stroke by stroke, text written stroke by stroke, locally at zero cost and re-rendered in seconds. Useful for teaching-video makers; the limit is it needs an agent to drive it, and complex layouts still need manual tuning. #Agent Skills#Animation#Video#Learning

    Whiteboard explainer animation skill. Turns an explainer script into a "real hand-drawing process" video — shapes drawn stroke by stroke, text written stroke by stroke, drawing wherever the narration …

    liangdabiaoPython★14

  • B runnel A small Go HTTP egress gateway with per-domain rate limiting, circuit breaking, queuing, caching and SSRF protection. A narrower option than Nginx for teams needing safe agent internet access; maturity and load-test data are not yet provided. #Network Security#APIs#DevOps

    A small, secure HTTP egress gateway with per-domain rate limiting, circuit breaking, queueing, caching, and SSRF protection.

    lynchestGo★11

  • B minidauth Decentralized authentication that solves the need for a key no one holds and governance capability in existing auth systems. It gained 94 stars in 1.4 days, but the use case is niche; commercialization means an identity layer for distributed systems or an open standard, but many cryptographic alternatives exist, security audits are hard, and user trust is difficult to build. #Secrets#Privacy#App Security

    Mini decentralised auth. Gives an existing auth system a key nobody holds, with governance over who can decrypt.

    sashyoJava★94

  • B memory-hub Local-first cross-agent working-memory retrieval on Windows, addressing a real pain point: multiple coding agents each remember only their own context. But mem0, basic-memory and many others already cover this space, and it doesn't explain its indexing approach or how it differs, so no advantage is visible yet. #AI Agents#Knowledge Base#Windows#Privacy

    Local-first cross-agent work memory search for Windows

    Dumpling-DragonPython★17

  • B codex-glass It targets a specific pain point: on Windows you can't see Codex's token consumption and quota, so it builds a floating window with local SQLite history merging. The scope is clear and it works out of the box; similar usage dashboards are common, and its difference is being desktop-resident and Chinese-oriented. #Windows#Productivity#Monitoring#codex

    A native Windows Codex usage floating widget and liquid glass dashboard: tokens, cost estimates, quotas, and SQLite history deduplication and merging.

    GGBond2424648901Python★14

  • B DiffExplorer A differential analysis tool for lab researchers who don't code, unzip-and-run, producing statistics and charts directly for group meetings, with a clear audience and goal. The worry is that no-code tools like this tend to be vague on statistical methods, so you need to see which experimental designs it supports. #Data Analysis#Science#Biotech

    Differential analysis tool for coding-free researchers – instant statistical results and charts, ready for your lab meeting. Just unzip and run! No-coding!

    JkingYuPython★27

  • B learn-from-materials Converts PDFs, books and papers into interactive learning web pages with traceability, quizzes and notes. It gained 59 stars in 1.7 days in the education AI space; commercialization means a paid tool or cloud service for students and enterprises, but similar products exist and conversion quality is hard to guarantee. #LLMs#RAG#Learning#Knowledge Base

    Turn PDFs, books and papers into interactive learning webpages｜Turn complex materials into traceable, quizzable, note-taking learning webpages

    dmoshehun-progPython★59

  • B ReUWP A compatibility layer that translates UWP Minecraft Bedrock into Win32, targeting only two old versions, 0.15.10 and 1.1.5. The technical substance is real and the scope clearly stated, making it a good sample for studying the UWP sandbox; but it's nearly useless to ordinary players, and playability is doubtful. #Game Porting#Reverse Engineering#Windows#minecraft

    ReUWP is a UWP to Win32 translation layer for Minecraft Bedrock 0.15.10/1.1.5

    OptiJuegosC++★11

  • B cli-tools Collects scattered Mac CLI tools into one directory and provides agent-friendly JSON output, useful for people who want local command-line tools callable by AI. But it doesn't explain how tools are discovered and indexed, and what the native app actually does is vague, so the value remains unproven. #CLI#AI Agents#macOS

    One catalog for every CLI tool on your Mac: native app plus a JSON-friendly CLI for AI agents

    flaviocopesSwift★21

  • B grove A Swift-written MCP server that lets AI assistants read and write macOS Reminders and Calendar directly, currently covering only these two native apps. Anyone using a Claude-style client on a Mac can install it immediately, and the native implementation is lighter than a Node bridge; the limitation is narrow scope, and several Apple calendar MCPs already exist. #MCP#macOS#AI Agents

    Native Mac tools for AI assistants. A Swift MCP server for Apple apps, starting with Reminders and Calendar.

    thsnkhnSwift★9

  • B novel-agent-studio A local desktop AI workbench for Chinese novel writing that separates story planning, chapter writing, evidence-based review, local revision and story memory, more concrete than the usual 'one-click novel generation'. But the field is crowded, the repo is only seven days old, and completeness and long-text consistency remain unclear. #LLMs#Content Writing#Knowledge Base

    A local desktop AI workbench for Chinese novel writing: story planning, chapter-by-chapter writing, evidence-based review, local revision, and story memory.

    skuramatataJavaScript★32

  • B supabase-multitenant This wraps self-hosted Supabase into a Supabase.com-style multi-tenant platform that runs on VPS, Coolify, or Dokploy, addressing the real need of giving each client an isolated instance. But the description is marketing-heavy, and whether isolation boundaries and upgrade paths are reliable can only be verified by reading code; for now it looks more like a set of orchestration scripts. #Databases#Self-Hosted#DevOps#supabase

    supabase-multitenant lets you self-host Supabase as your own Supabase.com-style cloud, running anywhere you choose—VPS, Coolify, Dokku, Dokploy, or your own infrastructure. Manage everything from one …

    supabase-multitenantTypeScript★20

  • B boss-job-agent A self-hosted job-hunting agent that runs workflows on BOSS直聘 through visible browser automation and leaves reviewable receipts. Batch applicants have a real need, and "visible browser plus receipts" is more honest than a pure script. The limitations are platform risk controls and account risk, and configurability is unverified. #Jobs#AI Agents#Automation

    Self-hosted Codex job-search agent with configurable workflows, visible browser automation and durable receipts

    xirichuyiJavaScript★14

  • B claude-code-statusline Adds a status line to Claude Code showing project, context usage, session duration, and quota, integrated with tmux. The need is real, but similar scripts are already plentiful, making this more a small config than a new tool. #CLI#Terminal#Productivity

    Claude Code status line: project, context window, session time and usage limits, with tmux-vitals integration

    k8adevShell★15

  • B agent_room A CLI for multiple people to collaborate on the same Codex session and project, with participants needing host approval; the idea is concrete and rare. But with only 9 stars and under two days old, it doesn't explain how multi-user conflicts are handled or how it works with git, so for now it's just an interesting prototype. #AI Agents#Collaboration#CLI

    CLI collaboration on one shared Codex session and project, with host-approved participants.

    DoorwoodGo★9

  • B chlorine-server A from-scratch inference engine rewrite for AMD Strix Halo that keeps the same wire protocol and OpenAI-compatible API, well-matched to the hardware and ambitious. But the upstream halogen 0.1.3 and the target models are hard to verify, and there's no evidence of it running, so it reads more like a statement for now. #LLMs#GPU#Performance#amd

    Clean-room AGPL-3.0 reimplementation of the halogen 0.1.3 inference engine for Qwen3.8-27B on AMD Strix Halo (gfx1151) — same wire protocol and OpenAI-compatible API. Not affiliated with Peonist.

    Heretek-AIHIP★9

  • B swimmich A native SwiftUI iOS client for Immich, aiming for feature parity with the official Flutter version. Attractive to self-hosted photo users, but claiming full parity after a day of building is unrealistic, and long-term maintenance is doubtful. #iOS#Self-Hosted#Images

    Native iOS SwiftUI client for Immich — self-hosted photo & video management. Full feature parity with the Flutter mobile app.

    millianlmxSwift★15

  • B ai-memory-skillpack A "bounded" project memory skill pack for Codex CLI and Claude Code, pitched at limiting memory size so it does not blow up the context. The direction is right, but this space is extremely crowded lately, and it does not explain retrieval, eviction, or conflict-handling strategies, making it hard to judge how it beats existing options. #Agent Skills#LLMs#AI Agents

    Bounded project-memory skills for Codex CLI and Claude Code.

    tudoumashuPython★22

  • B DuckDuckMove A Windows 11 avatar animation tool with a built-in duck GIF, support for custom animations, and one-click restoration of the original avatar; the need is small but clear, and the description is honest. The limitation is that it has only one feature and works by changing the system account picture, so whether it breaks after system updates is unknown. #Windows#Productivity#Images#Animation

    Make your avatar move. A Windows 11 animated avatar tool supporting the built-in duck GIF, custom GIFs, and avatar restoration.

    wooiC#★16

  • B DepthBenchCAD A three-tier evaluation for generative CAD systems, testing audit depth and evidence allocation, a new and narrow topic. The problem is it only matters to the small circle doing CAD generation research, and it doesn't say whether the eval set is public or how to reproduce it. #Benchmarks#ML Research#3D

    A three-level benchmark for evaluating audit depth, evidence allocation, and reliability in generative CAD systems.

    HongyeYangGTPython★13

  • B daro A lightweight desktop database client written in Flutter, explicitly pitched as a Navicat alternative. The direction is clear, but DBeaver and Beekeeper already dominate this lane, and with under a day of existence and 9 stars, it's unclear why anyone would switch. #Databases#Open-Source Alternatives#navicat

    A Navicat alternative; a lightweight desktop database management tool built with Flutter

    SpringHguiDart★9

  • B underplan A cable routing planner for Multiboard and Underware printed wall systems, a very narrow audience but a real need if it fits. The problem is it only serves this one hardware set, so anyone without these modules can't use it at all. #3D#Hardware#Productivity

    Interactive Underware & Multiboard Cable Management Planner

    francycalleTypeScript★16

  • B tagtinker-adv A small toolset for tinkering with electronic shelf labels on the Cardputer ADV; the hardware modification angle is interesting. But the description doesn't elaborate on ESL, with no wiring or protocol details, making it hard for outsiders to reproduce. #Hardware#Embedded#IoT

    A toolkit for Cardputer ADV to tinker with ESL's

    i12bp8C++★17

  • B midi-m8 A small command-line tool that converts Dirtywave M8 projects into MIDI tracks, a real need for M8 users. But the audience is tiny, it doesn't say which tracks or firmware versions are supported, and the same author publishing three repos in one day also lowers credibility. #CLI#Audio#midi

    A command line tool for turning Dirtywave M8 songs into Midi tracks

    ituitis-akyildizer16Language unknown★9

  • B cfmeeting A multi-platform meeting demo built on Cloudflare Realtime, covering macOS, Windows, iOS, Android and Web; open-source examples of such realtime capability are still rare, so the implementation is worth a look. But it calls itself a demo and lacks deployment docs and signaling notes, so it is far from usable as a product. #Streaming#Web Frontend#cloudflare

    An open-source multi-platform meeting app based on Cloudflare realtime-demo, supporting macOS, Windows, iOS, Android, web, and H5

    qianyubtcTypeScript★26

  • B architecture-assessment Turns architecture review into an agent skill, bundled with Roslyn syntax analysis and a SQLite source directory. The author has standing in the .NET community, but what problem the three pieces solve together and how to run them aren't explained, making it look like a personal workflow dump that outsiders can't easily use. #Agent Skills#Code Quality#AI Agents#roslyn

    Evidence-first architecture assessment and simplification skills, with Roslyn syntax analysis and a SQLite source catalog

    AarononthewebPython★13

  • B pc98-ai-localization-handbook A research handbook on localizing old PC-98 games, covering FM/SSG, MIDI and PCM verification flows, written in Traditional Chinese for easy AI reading. The material itself is scarce, but it's essentially a docs repo that can't be used directly, serving only the small retro-game translation crowd. #Documents#Learning#Audio#pc-98

    AI-readable PC-98 localization research handbook in Traditional Chinese, with FM/SSG, MIDI and PCM validation workflows and synthetic examples.

    BillyJr0127Python★23

  • B speechloom A research scaffold for speech models covering waveform encoding, causal decoding, codec tokens, and streaming, aimed at researchers building speech LLMs. But it lacks reproducible results and minimal examples, and gaining 30-plus stars within half a day of creation looks more like hype. #Speech#ML Research#LLMs

    End-to-end speech model research: waveform encoding, causal decoding, codec tokens, joint losses and streaming.

    fufuchiuPython★31

  • B viewport-ruler A Chrome/Edge extension that switches the viewport among five standard widths with one click, simulated in pure CSS and applied only to the current tab. Lightweight, but DevTools responsive mode already does the same thing; the only difference is fewer clicks. #Browser Extensions#Web Frontend#UI Design

    Chrome/Edge responsive debugging extension: click to switch the page viewport width (pure CSS simulation, controls only the current tab), with five built-in standard presets and a real-device checklis…

    kaythoneJavaScript★10

  • B midori-companion-avatar An unofficial fan project for 明前奶绿: a local web AI companion avatar with audio lip-sync, dynamic character art, lyric visuals, and generic Agent integration. Its positioning is honest and boundaries clear, but completeness is more like a fan toy, and the actual lip-sync and Agent integration need to be tried firsthand. #Animation#Speech#LLMs

    Unofficial, non-commercial fan project for Mingqian Nailv: a local web AI Companion Avatar supporting audio lip-sync, dynamic character art, floating icons, lyric visuals, and generic Agent integratio…

    tosya301JavaScript★21

  • B PassTheCert-rs A Rust rewrite of the existing PassTheCert, using Schannel to authenticate LDAP/S with certificates, handy for Windows internal network assessments. But it's just a port that solves no new problem and doesn't explain how it differs from the Python original. #App Security#Windows#Network Security

    Tool to authenticate to an LDAP/S server with a certificate through Schannel written in Rust. 🦀

    g0h4nRust★13

  • B ClayGlobe A SwiftUI clay-style 3D globe component that plants a waving flag in each country you've visited, suitable as UI material for travel apps. The craft is appealing but the use is single-purpose, and ready-made globe components are plentiful. #iOS#3D#UI Design

    A claymorphic 3D globe for SwiftUI that plants a waving flag on every country you visit

    haplollcSwift★17

  • B speechturn A Chinese-language research toolkit for speech LLMs, including audio instruction data, acoustic adapters, CPU micro-training, and evaluation. The direction is clear, but it lacks results and an onboarding path; it overlaps with speechloom, which appeared the same day, and the division of labor isn't clear. #Speech#ML Research#LLMs

    A speech large model research toolkit: audio instruction data, acoustic adapters, CPU micro training and inference, streaming events and evaluation.

    howervooPython★32

  • B netflix-tube It wraps YouTube in a Netflix-style browsing interface, and the author admits it was written casually out of boredom, which is honest. It's fun but has zero maintenance expectations, and similar frontend skins are plentiful, so it's worth a glance but not study time. #Video#Web Frontend#Streaming

    i was bored so quickly built this stupid tool that turns youtube into netflix.

    PankajtanwarbannaJavaScript★24

  • B mousiki A customizable TUI music player focused on being lightweight and fast. Mature terminal players like cmus and termusic already exist, and this offers no differentiating capability. #CLI#Terminal#Audio

    Highly customizable TUI music player designed to be fast and lightweight

    itzender5820C++★19

  • B claude-agent-sdk-bridge Explains how to hook subscription quotas into a self-built chat frontend using the Claude Agent SDK, with parameters, permission callbacks and deployment notes. Practical for those wanting to save API fees, but it is a tutorial note and the approach risks violating terms of service. #LLMs#AI Agents#Learning#SDKs

    Using the Claude Agent SDK to connect subscription quota and the full harness to a self-built chat frontend: differences from claude -p, query() parameters, permission callbacks, fixed prefix, prompt …

    marikaguraLanguage unknown★13

COrdinary58
  • C JustForPixel-ExpressiveLab An Android Material 3 motion and component library, now released as open source. It gained 39 stars in 1.6 days for mobile development; commercialization means a reusable UI library with community support, possibly commercial licensing or sponsorship, but Material 3 already has official components and the library needs ongoing maintenance to match API changes. #Android#UI Design#Animation#SDKs

    Publishable Open-Source Android Material 3 Expressive Component & Motion Library by Er. Mohd Amaan

    mohdamaan1Kotlin★39

  • C YourDesk A cross-platform remote desktop with hardware acceleration and multi-monitor support, competing with RustDesk and TeamViewer. It gained 72 stars in 1.8 days, but the market is crowded; commercialization means an enterprise or self-hosted edition, yet users prefer mature options, RustDesk is the open-source benchmark, and maintenance costs are high. #Networking#Self-Hosted#Windows#macOS

    YourDesk cross-platform remote desktop: supports macOS and Windows, hardware acceleration, multiple monitors, and clipboard transfer of text, images, files and folders.

    VaderChenGo★72

  • C SureForge An agent skill for complex work: research, ask, plan, build, verify and independently review, pure text with no runtime. It gained 46 stars in half a day, but it's essentially a checklist without technical moat; monetization is selling it as a prompt template with limited value, and the risk is being replaced by fuller agent frameworks or IDE integrations. #AI Agents#Agent Skills#Prompting#Productivity

    Agent Skill for complex work: research before asking, ask before planning, plan before building, verify before delivering, independent review before calling it done. Plain text, no runtime.

    Da7-TechPython★46

  • C TLPE Discloses a logic flaw in Android Telecom services that lets unauthorized apps execute code with system_server privileges. It gained 50 stars in 0.8 days and is pure security research with no commercialization path beyond defense; it loses value once patched, and exploiting it is illegal. #Exploits#App Security#Android

    CVE-2026-49881, a logic issue in the InCallController class in Android 17's Telecom service that allows an unprivileged app to gain arbitrary code execution as UID 1000 system_server

    SupersonicKotlin★50

  • C mactap-app MacTap triggers shortcuts by tapping a MacBook, a small macOS utility. It gained 230 stars in 1.0 days, notable attention, but it is a consumer tool with little practical value for developers; commercialization could be a paid app, but the single feature is easily absorbed by system gestures or third-party software, and it depends on hardware traits with a limited user base. #macOS#Automation#Productivity#Hardware

    MacTap — knock your MacBook to run shortcuts

    jaskirat1616Swift★230

  • C Learn-Rust-Attack-Vectors A Rust web3 attack-surface checklist for Solidity auditors, and the author is upfront about the target audience. But it's essentially a study-notes content repo, with no indication of how many runnable samples each vulnerability has or how it differs from existing audit handbooks. #App Security#Learning#Blockchain

    A hands-on catalog of Rust web3 vulnerability patterns, built for a Solidity auditor learning the Rust attack surface.

    IvanFitroRust★15

  • C agentic-stack-desktop A macOS local workbench aiming to unify memory across Claude Code, Codex, OpenCode, and Cursor via a knowledge graph. The pain point is real, but unified agent memory already has mem0, basic-memory, and others, and this offers only vision-level description without saying how the graph is built or conflicts handled. #AI Agents#macOS#Knowledge Base#LLMs

    Native macOS workspace with one local knowledge graph across Claude Code, Codex, OpenCode, and Cursor.

    codejunkie99Python★24

  • C jadense-in-zotero A Zotero plugin adding an AI research assistant, aimed at heavy literature users. The description is a single sentence with no mention of which model it uses or what it can do; Zotero already has several AI plugins. #AI#Plugins#Knowledge Base

    Jadense in Zotero — AI research assistant for Zotero

    jadense-aiTypeScript★10

  • C Tone3000-PresetBuilder Generates preset chains for Tone3000 neural amp models from artist, style, or track descriptions, aimed at guitar modeling users, and the idea is concrete. But it depends on Tone3000's preset format and API, the description shows no sample output, effectiveness cannot be judged, and the audience is very narrow. #Audio#AI Agents#tone3000

    AI Agent drive preset builder for Tone3000 - Neural Amp presets. Input artist/style/track details or any description and AI will build a chain preset for you

    tlennon-iePython★12

  • C vesper-street A small isometric survival-horror scene built with Three.js: a procedurally generated voxel flooded town, wet street reflections, and a flashlight as the only weapon, running directly in the browser with no external assets. The procedural generation and reflections are interesting, but it is more an atmosphere experiment with little information on gameplay or playability. #Games#3D#Web Frontend#Game Dev

    Isometric survival horror in Three.js: a drowned town built from procedural voxels, wet-street reflections, a flashlight that is your only weapon. No assets, runs in the browser.

    winchxyzJavaScript★9

  • C swiftui-sky Procedurally generates sky backgrounds with clouds and seasonal lighting in SwiftUI, with iOS/macOS demos. It gained 79 stars in 1.4 days, but it is a visual spectacle with limited value to developers beyond aesthetics; there is no commercialization path beyond a component library, and it risks being replaced by more mature design libraries or system APIs. #iOS#macOS#Animation#UI Design

    A procedural SwiftUI sky with drifting clouds, seasonal light, stars, and meteors. Includes an interactive iOS and macOS demo.

    zolplay-labsSwift★79

  • C CVE-2026-83991-writeup-and-poc A writeup and PoC for bypassing Windows cloud file access checks, useful for Windows sandbox and local privilege assessments. But the content is thin, lacking impact analysis and remediation advice, and the numbering and details need verification. #Exploits#App Security#Windows

    CVE-2026-83991: Windows Cloud Files access-check bypass

    karolloooolC★11

  • C ref2va-h3-video-optimizer A prompt iteration optimizer for MiniMax-H3 reference-to-video generation, requiring local ComfyUI plus an LLM to run. The sequential iterative feedback idea is decent, but it's locked to a single model, and results can't be verified outside that environment. #Prompting#Video#LLMs

    MiniMax-H3 Reference-to-Video prompt optimizer (v1.0) — uses a sequential iterative feedback method to automatically optimize a video's prompt to a higher-quality version; requires local ComfyUI + a l…

    klfzqxsPython★14

  • C OmniStudio A local LLM desktop workbench bundling model downloads, three inference engines, and chat, voice, image and OCR apps. It aims to swallow LM Studio and Open WebUI, but a shell built in one day is unlikely to hold up on stability. #LLMs#Self-Hosted#Speech

    OmniStudio is an all-in-one local large model desktop workbench (Electrobun + React), combining a model marketplace for downloads, inference management for the three engines llama.cpp/vLLM/SGLang, and…

    kunpengtalkTypeScript★12

  • C antigravity-account-switcher A small macOS tool for one-click switching between multiple Google accounts in Google Antigravity. The need comes from free quotas and it essentially circumvents the account system; single-purpose and easily killed by official changes. #macOS#Productivity#Automation

    Fast 1-click Google account switcher for Google Antigravity on macOS

    m4tinbeigi-officialPython★8

  • C fleet-skill A bounded multi-agent scheduling Skill for Codex that emphasizes task-based selection, explicit write permissions, and convergence acceptance, more concrete than most 'agent orchestration' slogans. But the description is a single Chinese sentence with no example tasks or scheduling strategy details, so outsiders can't judge whether it actually works; noted but not listed. #AI Agents#Agent Skills#LLMs

    Fleet 星河舰队: a bounded multi-Agent scheduling Skill for Codex, selecting by task, explicitly defining write permissions, and converging for acceptance.

    WoltziaraLanguage unknown★10

  • C opencode-plugin-lane A copy-on-write worktree plugin for OpenCode V2 that gives each parallel agent its own lane, a targeted idea rather than a generic orchestration layer. But it only serves OpenCode users, has 8 stars and no usage examples, so stability remains unclear. #AI Agents#Plugins#Editors#opencode

    Lane copy-on-write worktree strategy for OpenCode V2

    anomalycoTypeScript★8

  • C technical-image-generation-skill An Agent Skill for technical illustrations, from an author with a reputation in CSS circles, on a solid topic. But the repo has only a one-line description, not saying what the input is, what kinds of images it produces, or which model it depends on, so others can't reproduce it directly. #Agent Skills#Image Generation#AI#Images

    High-quality technical illustration SKILL

    chokcocoLanguage unknown★12

  • C infinite-sugar A browser-based fruit fly connectome artwork that turns connectome data into browsable visuals, interesting in both subject and presentation. But it's an artwork, not a tool, with no reusable value for developers' daily work, better saved to a design inspiration folder. #3D#Biotech#Science#Web Frontend

    Infinite Sugar — a browser-based fruit fly connectome artwork

    cnqsoTypeScript★17

  • C shittim-logon A proof of concept that replaces the Windows 11 login screen with a Spine real-time rendered scene; the author states there are no assets and it can't be built. The idea is interesting and the honesty is commendable, but it's completely unusable for anyone wanting to try it. #Windows#Animation#Games

    Replacing the Windows 11 logon screen with a live-rendered Spine scene: the concept and the adapter. No game assets, no Spine runtime -- it does not build.

    helloyorkC++★15

  • C pawbot A self-hosted AI agent with WebUI, CLI, MCP tools and record-replay debugging. Record-replay is a small highlight, but overall it is another feature-list agent framework in a crowded field. #AI Agents#MCP#Self-Hosted

    A self-hosted AI agent with WebUI, CLI, MCP tools, multi-provider support, and Record & Replay for deterministic debugging.

    m2dumplingPython★9

  • C PentestPi Self-described as an autonomous penetration-testing agent framework with external state memory and guardrails against false conclusions. The direction is reasonable, since pentest agents most fear false vulnerability reports, but the description gives no scope, model dependencies, or test results, and similar autonomous pentest projects are already numerous. #App Security#AI Agents#Network Security#Exploits

    Autonomous Penetration Testing Agent Framework with External State Memory & Anti-False-Falsification Guardrails.

    weidutechPython★10

  • C awesome-competitive-olympiad-algorithms A collection of competitive programming problems and lecture notes spanning NOIP to ICIP level. It gained 65 stars in 1.7 days, but it only organizes resources and creates no new tool; commercialization is slim, at best teaching material, and many similar repos exist with quality varying by individual maintenance. #Learning#Knowledge Base#noip

    Original algorithmic contest problems and lecture notes — NOIP to NOI & ACM-ICPC level

    wzj52501C++★65

  • C loopera It bills itself as a hypothesis-driven factor research agent with evidence gating and research memory. Sounds good, but it's all research-paradigm description with no explanation of what data it connects to, how factors are computed, or how results are validated, so finance practitioners can't use it today. #AI Agents#Fintech#Data Analysis

    A hypothesis-driven AI agent for fundamental factor research, with evidence-gated validation and research memory.

    Loopera-aiLanguage unknown★24

  • C ai-motion-poster-skill An agent skill pack for dynamic posters, with four-act storyboards, object-driven transitions, and prompt templates. It's essentially a prompt collection whose output can't be verified, and similar skill repositories are appearing in bulk. #Agent Skills#Prompting#Image Generation#Content Writing

    High-end commercial motion poster Skill: photorealistic 3D objects × experimental typography, four-act storyboard, object-driven transitions, accurate copy and reusable prompts.

    huangbai-AILanguage unknown★13

  • C fly-with-me A browser toy where a bird flies over an infinitely generated landscape. Fun as a demo, but it solves no concrete problem and leaves nothing reusable behind. #Web Frontend#Animation#Games

    Fly with me - watch a bird fly across an infinite gorgeous world.

    kunchenguidJavaScript★17

  • C gpt-images A Claude skill that expands one requirement into GPT Image six-part prompts, with notes on API limits. Essentially a prompt template pack, and such content is everywhere; value depends entirely on template quality, not engineering. #Prompting#Image Generation#Agent Skills

    A Claude skill that turns a plain brief into a production-ready prompt for OpenAI GPT Image models (ChatGPT Images 2.5): six-block prompts, the hard API limits, and edits that don't drift. MIT.

    lukasersilLanguage unknown★9

  • C jobscraper A scraper that grabs job postings by role keywords and pushes them to you — a real need, but similar projects and existing aggregators are plentiful. It doesn't say which sites it supports, how to configure it, or how long it runs stably, so there's no reason to switch from job platforms. #Jobs#Automation#Email

    A scraper with crawler to find and send you the newest and most suitable jobs for your current job title

    AhmedAbdelgelelJavaScript★16

  • C Turbo-IO An unofficial SDK and iOS client for RayNeo iO glasses enabling voice AI and other features. It gained 55 stars in 1.5 days but depends on niche hardware; commercialization is limited and the official SDK may replace it, with uncertain hardware sales and third-party compatibility and maintenance burdens. #SDKs#iOS#Hardware#Speech

    Unofficial SDK and iOS client for RayNeo iO: non-commercial study and research, voice AI, recording, custom notifications, Codex, and Web display preview. BES2800 lightweight glasses, Android pending …

    Turbo1123Swift★55

  • C takeoff-launcher A Raycast-inspired native Windows launcher in C++ focused on fast search and launching system tools, with a clear direction. But the space is crowded, with Flow Launcher and PowerToys Run both open source and mature, and the repo doesn't say what it does better. #Windows#Productivity#Open-Source Alternatives#raycast

    A fast, native Windows productivity launcher inspired by Raycast. Search and launch apps, system tools instantly.

    akiraengC++★9

  • C vision-demos A collection of 'fun' computer vision examples, fine for classroom demos. But as a repository it has no reusable components and lacks a unified entry point or documentation. #ML Research#Images#Learning

    Fun real-world computer vision demos!

    jeremyiparkPython★29

  • C agy-context-monitor A small ring showing context window usage for the Antigravity desktop 2.0 client, copying the form of Codex's equivalent. It is a handy widget for heavy Antigravity users, but it has a single function, is tied to a specific client version, and similar usage monitors already exist in editor plugins. #Monitoring#Productivity#antigravity

    A context window monitoring ring for the antigravity desktop 2.0 (same as Codex)

    xiaoliulian99JavaScript★10

  • C opencode-go-proxy-for-trae A local reverse proxy written to fix the 400 error when Trae connects to OpenCode Go, injecting session headers and rewriting model names. The problem is specific and the description honest, but it's just a temporary patch for two closed-source services that breaks as soon as upstream changes. #APIs#Networking#Automation

    Local reverse proxy: bidirectional forwarding between Trae and OpenCode Go (Zen API), automatically injecting the session/request headers required by upstream (`x-opencode-session`, etc.) and rewritin…

    LIMTCYTPython★9

  • C virtual-ai-infra-team Claims to be a virtual AI Infra team that automatically discovers, stress-tests, and safely upgrades local inference, with quality gates and rollback. The vision is enormous, but there's no concrete entry point or reproducible results, and this kind of fully automated ops narrative needs evidence most. #AI Agents#DevOps#LLMs

    An open-source virtual AI Infra team that discovers, benchmarks, and safely upgrades local LLM inference—with independent quality gates, a stable OpenAI-compatible API, and automatic rollback.

    hsj576Python★11

  • C astra-motion-recreation-remotion A Remotion recreation of an existing motion design, with editable source, assets and reproducible rendering, plus a comparison to the original. It has some reference value for people learning Remotion animation, but it's essentially copying someone else's work, raising no new questions and reading more like tutorial material than a tool. #Animation#Video#Web Frontend#remotion

    Astra Motion Recreation — Remotion. Editable animation source, bundled assets, reproducible renders, and a comparison with the original by @rajzmotion.

    TejashmakwanaJavaScript★25

  • C catalog A list of Docker Compose recipes for self-hosted apps, claimed to be reviewed, saving self-hosters from digging through official docs. It's essentially a resource-list repository whose value depends entirely on review quality, which can't be judged yet. #Self-Hosted#DevOps#Linux

    Public catalog of self-hosted applications with reviewed Docker Compose recipes

    fossaryhqShell★20

  • C icebreaker A local AI chat app built with Rust and iced on top of Hugging Face and llama.cpp, with a description full of emoji. It piles on tech labels without saying how it beats existing local chat frontends—a typical yet another wrapper. #LLMs#Self-Hosted#GPU

    A local AI chat app powered by 🦀 Rust, 🧊 iced, 🤗 Hugging Face, and 🦙 llama.cpp

    ituitis-akyildizer16Language unknown★12

  • C html-visualizer A Skill that has AI render long replies as web pages, compatible with skill-supporting tools like Claude Code, Codex, and Cursor. The idea is practical but very light, and after reading you still don't know how styling is controlled or whether output is stable; similar markdown-rendering and page-generation options already abound. #Agent Skills#AI#Web Frontend#Documents

    Let AI turn long replies into beautiful, readable web pages — for AI tools that support skills, such as Claude Code, Codex, Cursor, Cline, Copilot and more

    chenjackle45HTML★9

  • C GitCard Generates GitHub social preview images in the browser with local 2x PNG export, a real need with a simple implementation path. But mature tools like socialify already fill this gap, and what's extra here is just aesthetic templates and pure frontend; 9 stars show no differentiation. #Images#Web Frontend#UI Design#Social

    ⚡ An exclusive social preview image generator for open source projects | Aesthetic GitHub social preview card generator, 100% in-browser, 2x PNG export.

    chensy787TypeScript★9

  • C awesome-astra-blender-characters An agent skill for Blender character creation and repair, with nine-view references and a face and hair workflow. Named 'awesome' but essentially a process template and prompt collection, lacking verifiable output. #Agent Skills#3D#Prompting

    An agent skill for Blender character creation and repair, with nine-view references, face and hair workflows, and visual validation.

    icesixgodPython★12

  • C discover-unknowns It claims to help agents find blind spots and turn uncertainty into verifiable actions, but the whole thing is methodology slogans with no explanation of the skill's inputs, steps, or output format. Requirement-clarification prompts like this are already plentiful, and it offers no actionable difference. #Agent Skills#Prompting#AI Agents

    Discover blind spots, surface implicit requirements, and turn uncertainty into verifiable action. An Agent Skill for requirement clarification, FDE delivery and development acceptance.

    EscapeVelocity-AILanguage unknown★11

  • C zhiwen-wo A skill that turns ideas into a structured brief and execution prompts through follow-up questions. Requirement-interview prompts like this are already common, and it gives no process details or differences from similar skills, just a Chinese wrapper. #Agent Skills#Prompting#Productivity

    An adaptive requirements interview skill that turns ideas into a structured brief and an execution-ready prompt.

    Mortis114114Language unknown★10

  • C MusicPlaygournd A real-time music editor for macOS written in Swift, but beyond one line there's no usage, no screenshots, and even the repo name is misspelled "Playgournd." Who it's for, what it edits, and how to start are all blank, so it can't be judged worth installing. #Audio#macOS#Editors

    A native macOS live music editor powered by SwiftMusic

    1amageekSwift★17

  • C Chinese-IP-Law-Basic-Rule-Book- Chinese IP laws and cases compiled in Markdown with commentary, for easy AI reading and Obsidian browsing. The organization shows care, but it's still a text reference repo limited to legal practitioners, with no direct use for developers. #Documents#Knowledge Base#Learning

    Basic laws and regulations of Chinese intellectual property law + judicial interpretations + included cases + selected answers from the Legal Q&A Network, compiled in commentary style, in md format su…

    StefanCHEN2026Language unknown★8

  • C tangle A web app claiming to build machine learning pipelines without environment setup. There are no technical details or examples, and its features heavily overlap with existing visual pipeline tools, so it's impossible to tell whether it actually runs. #ML Research#Web Frontend#Automation

    Tangle is a web app that allows the users to build and run Machine Learning pipelines without having to set up development environment.

    ituitis-akyildizer16Language unknown★10

  • C world-radio A 3D globe radio built with Vue where clicking a location plays local stations; a likable little toy. Technically about three.js plus radio streams, with no word on station data sources, availability or how to run it; a one-off small project. #Audio#3D#Web Frontend

    3D Earth radio website

    ns2250225Vue★15

  • C OHnews A multi-source news terminal for cognitive-loop analysis, using LangGraph for structured extraction and evidence-backed judgments. There are more buzzwords than features, with no explanation of sources, criteria, or what the output looks like, so no gap versus ordinary news summarizers is visible. #AI#LLMs#AI Agents

    AI-native news intelligence terminal: cognitive-loop analysis over multi-source news with LangGraph agents, structured information extraction, and evidence-anchored judgments

    OrangeLattePython★13

  • C astrbot_plugin_what2play An AstrBot plugin that recommends anime and galgames by adjusting probabilities based on ratings, popularity and tags. A niche entertainment plugin with no general value. #Plugins#Games#AI

    An AstrBot plugin that recommends anime and galgames, supporting modifying recommendation probability by rating and popularity, and recommending by tag

    ma114514Python★9

  • C astudio-wysiwyg-html-editor Yet another embeddable WYSIWYG HTML editor, with a business model written as "free basic version + paid updates," and copy that reads more like a product page than an open-source intro. The space already has mature options like Tiptap and Quill, and no added value is visible. #Editors#Web Frontend#Documents

    Free forever base version of Astudio WYSIWYG Rich HTML Editor for websites and CMS platforms. Optional paid updates and support by Astudio LLC.

    AstudioLLCJavaScript★11

  • C ComputerUse A small project aiming to hook computer use into tools like OMP and OpenCode, riding a hot direction. But the one-line description doesn't say what it wraps or how to integrate, and 11 stars prove no usability, so it's unclear it beats calling each vendor's agent directly. #AI Agents#Automation#Editors

    Use computer use with tools like OMP or OpenCode.

    ashhartJavaScript★11

  • C copy A copy-trading terminal on Robinhood Chain with token discovery, copy wallets and simulated matching. Crypto copy trading is inherently very risky, and the description is full of marketing words; not recommended for anyone. #Crypto#Blockchain#CLI

    Working CLI-first copytrade terminal for Robinhood Chain — live token discovery, profit hunters, COPY wallets, paper engine and native queue.

    lunarresearcherJavaScript★21

  • C SR-Paper-List A Science Robotics robot-learning paper list with cover verification and a searchable catalog, curated with care. But it's still a content collection whose value depends on long-term maintenance; readers are better off going straight to the journal or existing awesome lists, so it isn't worth opening separately today. #Robotics#Science#Knowledge Base

    Robot learning papers in Science Robotics (2024–2026), with verified cover papers and a searchable directory

    YanjieZeTypeScript★18

  • C Monozukuri A so-called skill claiming to inject Japanese craftsmanship into AI coding, with only philosophy, no executable method, and no examples showing improved output. A vague vision-type skill; a markdown. #Agent Skills#Prompting#Code Quality

    An AI coding skill that applies Japanese-inspired craftsmanship to build simple, correct, verifiable, and maintainable software.

    kuya-eggLanguage unknown★10

  • C memochka-bot A Russian-language AI bot for Telegram group chats with memory, local stickers and a fixed persona, aimed at people wanting a 'group member with personality'. Such bot templates are abundant, the description is all selling-point phrases and emoji, and deployment and memory storage are unexplained; just another group chat bot for now. #LLMs#Social#telegram

    😼 An AI participant for Telegram chats with memory, local memes, reactions, and its own personality

    murzichek5-specPython★20

  • C Wake-Trace An autonomous wake-up runtime for relational AI companions, concept-first, with no explanation of where it runs, how it triggers, or how it differs from existing agent scheduling stacks. The audience is niche and the need vague, so no usable information can be read from it yet. #AI#AI Agents#LLMs

    A lightweight continuous autonomous wake-up runtime for relational AI companions

    Maymay05072Python★12

  • C abstract-brawl A JS game cramming side-scrolling fighting, turn-based combat, and auto-battler into one project, self-described as non-commercial learning only. The scope is too broad with no design doc, making it a personal exercise; for open-source fighting games, the MUGEN ecosystem and existing engines are far more viable. #Games#Game Dev#Web Frontend

    Abstract Brawl: side-scrolling fighting, turn-based battles, auto chess. Open source, for non-commercial learning and exchange only.

    XzJoshZJavaScript★20

  • C Under-water-image-enhancement Stitches paper methods for underwater image restoration (dark channel, guided filtering, MSRCR and others) into a MATLAB implementation, a reference for students running restoration experiments. But underwater enhancement already has plenty of similar code and toolboxes, and with no sample data, result comparisons or run instructions, it is unclear how it beats existing implementations. #Images#Science#ML Research

    Underwater images suffer from color distortion, haze, scattering, and uneven lighting. This paper proposes a MATLAB-based restoration method using constrained color compensation, dark channel prior, g…

    montra1476-jpgMATLAB★12

  • C experience-finder A Python tool that scrapes Xiaohongshu content to compile interview experience, aimed at interview preppers; a real need. But similar scrapers and aggregators are many, and it depends heavily on platform anti-scraping and account status, so long-term availability is doubtful; the one-line description says nothing about cleaning, deduplication or structuring. #Jobs#Automation#xiaohongshu

    Organizing interview experience wins via Xiaohongshu

    songzhufuPython★12

DSkip97
  • D vintage-latex A LaTeX example collection with retro academic-paper aesthetics, including astronomy charts and star maps. It gained 125 stars in 0.4 days for its distinctive visuals, but it's essentially a template set that solves no real problem; monetization is selling templates as one-off purchases, with the risk of lacking long-term maintenance value and being replaced by modern typesetting tools. #Documents#Science#Learning

    Twenty LuaLaTeX and MetaPost (fiziko) examples of the old scientific-paper aesthetic: engraved plates, star charts, computed log tables, knots, telescopes

    FoadsfPowerShell★125

  • D truanayangi A Vietnamese lunch randomizer app with an unboxing animation, a casual project. It gained 301 stars in 1.6 days, showing popularity, but it has no technical moat, little practical value for developers and no commercialization path; hype fades fast, its lifecycle is short, and it cannot become a dependency. #Games#Productivity

    CS:GO-style lunch roulette. Mở hòm, chốt món ăn trưa.

    nagisanzeninTypeScript★301

  • D text-humanizer Rewrites AI-generated text to sound human so it bypasses detectors like Turnitin or GPTZero. It gained 65 stars in 1.4 days, but the use case is gray and amounts to a cheating tool, with poor commercialization prospects and compliance risk; detector upgrades and legal and ethical issues threaten it. #LLMs#Content Writing#Crack

    text-humanizer is an open-source project designed to convert text generated by AI to its "humanized" version. Bypasses the most of AI detectors such as Turnitin or GPTZero

    SpaceDudemPython★65

  • D job-search-tracker A job application tracker with location grouping and statistics, plus an overlay assistant and browser extension crammed in. The feature boundaries are muddled, the overlay's purpose is unclear, and similar spreadsheet templates are everywhere. #Jobs#Productivity#Browser Extensions

    Job search tracker with location-buckets, stats, overlay helper software and fastrack extension

    harisenthil1Python★18

  • D agentic-workflow-playbook A Korean-language workshop deck on agentic workflows, a content-type repository. For readers it's secondhand retelling, with nothing installable and no way to verify the methods. #AI Agents#Learning#Prompting

    Workflow and verification — an agentic workflow playbook (Samsung Electronics DX seminar secondary material)

    jha0313HTML★18

  • D x-bookmarks-to-obsidian Judging by the name, this is a script that exports X bookmarks to Obsidian, a real need for heavy X users who collect material. But the repository has no description, so it is impossible to confirm whether it uses the API or scrapes pages, which fields it keeps, or whether it syncs incrementally, and no usability evidence is available. #Productivity#Knowledge Base#obsidian

    sunkeke-aiPython★17

  • D reading-room Only one line, "give your favorite books a cozy home," without clarifying whether it is a reader, library manager, or bookmark tool. The author's past projects have a good reputation, but at this stage the repository gives readers no way to judge whether it is usable; atmosphere comes first and documentation is missing. #Documents#Knowledge Base#Web Frontend

    A cozy home for your favorite books.

    transitive-bullshitTypeScript★10

  • D ddalkkak-threads-community A community-edition Windows tool for Threads content production and automation, essentially a social media bulk-posting pipeline. Such scripts are everywhere, and the real variable is platform risk control rather than code quality, leaving nothing technically instructive for developers. #Marketing#Social#Automation#Windows

    Threads content creation and automation tool for Windows Community Edition

    apache3563-bitTypeScript★33

  • D fly-effect Claims to be 'resource-aware open research' on fruit fly neural circuits and a simulated organism, an ambitious framing. But the repo gives no concrete model, data, or experimental setup, making it more a direction statement than the connectome visualization projects on the same subject. #Science#Biotech#ML Research

    Resource-aware open research on neural circuits, bodies, and simulated organisms. Starting with Drosophila.

    dj-thankPython★10

  • D Jsbox-Ipa A small script for installing IPA files on iOS via Jsbox, with narrow, sideload-oriented use and almost no documentation. There is little engineering content for general developers to learn from. #iOS#Automation#Downloaders

    Jsbox install IPA

    domplingJavaScript★21

  • D iso-27001-guide A compilation of ISO 27001 clauses, Annex A controls, checklists, and templates that saves compliance people some lookup work. But it's still a document collection, less authoritative than the standard text and formal consulting materials, and contains no runnable tools. #Learning#Documents#App Security

    ISO/IEC 27001 practical guide. Requirements, Annex A controls, checklist, ISMS templates.

    easy-standartLanguage unknown★8

  • D claude-outreach A B2B outreach skill pack for Claude Code, with 16 sub-skills spanning lead finding, email lookup, and GDPR compliance. Feature-packed but essentially a cold-email mass-sending pipeline, with high maintenance and anti-spam risk—not recommended as a tool to bookmark. #Agent Skills#Marketing#Email#AI Agents

    Universal B2B outreach skill for Claude Code. 16 sub-skills + 5 sub-agents: ICP targeting, lead sourcing, email finder, phone finder, verification, AI personalization, multichannel sequences, delivera…

    emelia-ioPython★12

  • D nextjs-fullstack-saas-starter Yet another Next.js + MongoDB + shadcn SaaS boilerplate, described only with empty phrases like 'solid foundation for scalable SaaS', without saying what it adds over templates like create-t3-app. Not worth ten minutes unless billing and auth are actually runnable. #Web Frontend#Databases#Build

    A modern, open-source full-stack SaaS starter built with Next.js, TypeScript, MongoDB, and shadcn/ui a solid foundation for building scalable SaaS applications.

    mohamed-elsayed-200TypeScript★17

  • D alanstudio-framework A Portuguese thin harness for 'creating and iterating games with AI' that stresses it is 'not an engine', but beyond that positioning says nothing about what it offers or how it fits existing workflows. Little usable information for indie game developers, and no clear difference from several experiments stuffing LLMs into level generation. #Game Dev#AI Agents#Automation

    Alan Studios Framework — harness thin para criar e evoluir games com IA. Não é um motor.

    oalanicolasPython★9

  • D eidra-agent Crams persona agents, explicit memory, retrieval, adversarial smoke tests, and a post-training roadmap into one repository, an overly broad positioning. It may inspire people building persona agents, but every part stays at the idea stage and lacks a minimal runnable example, so it does not look usable today. #AI Agents#RAG#LLMs

    Persona agents with explicit memory, retrieval, adversarial smoke tests and a practical post-training roadmap.

    vigorleeJavaScript★9

  • D mi-fitness An archived mirror of the Mi Fitness family data SDK, querying heart rate, sleep, and steps through authorization. It has reference value for people wanting to integrate health data, but it is explicitly a mirror with uncertain future maintenance; since it involves personal health data, check the authorization scope and compliance boundaries before use. #Health#APIs#Privacy#Android

    Archived mirror of the Mi Fitness family data SDK (mi-fitness): query heart rate, sleep, steps, and other health data with family authorization

    wenxixixi-miaoPython★12

  • D content-forecast A content-creation skill claiming to handle topic selection, script review, shooting guidance and view-count prediction. The scope is far too broad, "predicting views" is nearly unverifiable, and input/output formats aren't specified — a typical AI-wrapper vision. #AI#Content Writing#Marketing

    A creator-first AI skill for topic generation, script review, filming guidance, and view forecasting.

    ColinjqqPython★15

  • D DLSS-5-MANAGER From the name it looks like a small tool for managing DLSS versions or configs, but the repo has no text description, so it's impossible to confirm what it does or which game files it replaces. A tool that drops DLLs into game directories should explain itself most clearly, and this gap is fatal. #Games#Windows#GPU

    NOD1X-TECHF#★12

  • D CCNA-PacketTracerlabs A collection of completed Cisco Packet Tracer labs that can save CCNA candidates time building topologies. But it's a material repo rather than a tool, and less systematic and authoritative than official courses and NetAcad labs. #Learning#Networking#Knowledge Base

    This repository contains a collection of completed Cisco Packet Tracer labs covering various topics from the CCNA (Cisco Certified Network Associate) curriculum. The labs provide hands-on practice for…

    abderaoufsecLanguage unknown★20

  • D LiteBSD A self-built Unix-like OS whose description is just 'that i built,' giving no sense of target platform or differentiation. A personal OS exercise; those wanting to read a kernel are better off with fully documented and tested SerenityOS or xv6. #Operating Systems#Linux#Embedded

    Unix-like Operating System that i built

    ahmedbarakat207C★15

  • D sharp "Control your social media feed" is just a slogan, with no indication of which platforms are supported, how filtering works, or whether it is a browser extension or client. A typical vision placeholder project that cannot be evaluated or used today. #Social#Productivity

    Control your social media feed, not the other way around.

    tshmieldevTypeScript★18

  • D prod-UAE-Corporate-Tax-Skill It got 78 stars in 0.6 days, but the repo has no description, so there's no way to tell whether it's organized rules or a runnable skill. UAE corporate tax is an extremely narrow audience, and no explanation plus abnormally high stars makes it suspicious for now. #Fintech#Documents#Agent Skills

    Git-AgniLanguage unknown★78

  • D awesome-rsi A list of papers and projects on recursive self-improvement, a typical awesome-style resource collection. The topic is interesting, but its value depends on the maintainer's curation standards, and lists like this aren't new tools in themselves. #Learning#ML Research#AI Agents

    A curated collection of papers, research blogs, and open-source projects on recursive self-improvement (RSI), self-modifying agents, and iterative self-training.

    PicrewPython★8

  • D resume-portfolio-template A resume and portfolio template with no description. Such template repos are everywhere, and 52 stars within a day of creation isn't reliable evidence; skip. #Web Frontend#UI Design#Documents

    christiezhou1215-hueHTML★52

  • D scry A self-hosted game and desktop streaming platform whose description claims 'DLSS 5 Neural Rendering' support, when DLSS 5 doesn't exist yet—that alone makes the description untrustworthy. Sunshine plus Moonlight already covers streaming, so there's no reason to look at this. #Streaming#Games#Self-Hosted#Video

    Scry is a self-hosted game and desktop streaming platform with support for DLSS 5 Neural Rendering.

    elowbePython★10

  • D ForgeFlow-Open-source-automation-tool-for-batch-file-media-processing-with-a-native-Windows-GUI Cramming 'open source, batch processing, native Windows GUI' into the repo name is typical keyword stuffing. Batch transcoding and media processing already have mature options like HandBrake and FFmpeg wrappers, no differentiation is given, and the description started rolling out within half a day of creation. #Windows#Automation#Video

    ForgeFlow Open-source automation tool for batch file/media processing with a native Windows GUI.

    bduxhlll12Go★12

  • D kakeya-3d A project formalizing the three-dimensional Kakeya conjecture in Lean; the topic is worthwhile and formalized math has advanced a lot recently. But the repo has no description, so it is impossible to tell how far the proof has progressed or whether it compiles; to anyone outside formalization peers, it is an empty repo. #Science#Learning

    project-numinaLean★22

  • D awesome-gpt-image-prompt-2.5 A collection of 2,383 GPT-Image-2.5 prompts packaged as a Claude Code plugin. It is essentially a content repository whose value depends entirely on individual prompt quality rather than code, and dozens of such prompt packs appear weekly, leaving little engineering substance to evaluate beyond a few usable templates. #Prompting#Image Generation#AI

    Prompt as Spec — 2,383 GPT-Image-2.5 prompts, each a complete design specification. Installable as a Claude Code plugin. By Devault.

    stretchcloudJavaScript★11

  • D 2026 A repository accompanying an EPFL course, containing 2026 course code and assignment materials, with no reusable tools for developers outside the course. It's coursework, so it gets a low score across the board. #Learning#Data Analysis#Science

    epfl-adaJupyter Notebook★9

  • D polox_ai A self-described agent-native multimodal generation platform that puts all interaction into chat and an infinite canvas, and invokes an unsourced concept called 'DeepSeek Harness'. With no verifiable usage or examples, it is a typical vague big-platform narrative. #AI Agents#Multimodal#Image Generation

    An open-source, agent-native platform for multimodal AI generation, built on DeepSeek Harness. Every interaction lives in agent conversations and an infinite canvas, unifying creation, generation, and…

    saihhold-zhaoTypeScript★19

  • D recipe-lab-sony-a6000 Judging by the name, this is a picture profile recipe tool for the Sony a6000, but the repository has no description or screenshots, so it is impossible to confirm whether it is a parameter calculator, recipe manager, or collection of straight-out-of-camera samples. There is currently no usable information for photography users. #Images#UI Design#sony

    voxivoidJava★14

  • D TechIstanbul_AI_Python A collection of AI/ML teaching code aimed at classroom students. It's a tutorial-notes repository with no usable new tools or unique methods for readers of the list. #Learning#ML Research#Data Analysis

    for ML,DL,AI learning

    hamitmizrakPython★20

  • D solvr-python An official SDK wrapping Kasada and DataDome bypass interfaces, with PoW minting thrown in. There's nothing technically new, and in nature it helps people evade anti-scraping controls, so it shouldn't be recommended on any list. #APIs#SDKs#Scams

    The official Python client for the Solvr API: typed sync and async access to the Kasada and DataDome endpoints, with free proof-of-work minting and full examples.

    Solvr-ioPython★12

  • D solvr-rust A Rust client for the same service, and changing the language doesn't change its nature: the core selling point is still bypassing Kasada and DataDome risk controls. Type safety is just packaging, and it's not worth recommending given its purpose. #APIs#SDKs#Scams

    The official Rust client for the Solvr API: async, fully typed access to the Kasada and DataDome endpoints, built on reqwest with zero runtime overhead.

    Solvr-ioRust★17

  • D pi-explabs The description is only 'Pi extension for Experiential Labs', with no indication of what it does, who it's for, or how to run it. A repository with no stated purpose. #Plugins#AI#Automation

    Pi extension for Experiential Labs

    joeljucaTypeScript★9

  • D tui Nothing but a repo name, so it's impossible to tell whether it's a TUI framework, component library, or CLI app. Getting 56 stars in a day makes an explanation more necessary, not less; a repo with zero information gets no space. #Terminal#CLI#Editors

    ZenNotesGo★56

  • D map-yue2.github.io A static website page for the YuE2 project, with no usable code or tools beyond presentation. #Web Frontend#Documents#Learning

    YuE2 project website

    map-yue2JavaScript★12

  • D fourthcross.tech The source code of a company's static website, which is website content rather than a reusable tool. It offers some reference for front-end developers wanting to imitate its page structure, but has little usable value for most developers. #Web Frontend#Marketing

    fourthcross.tech static website

    zerodhaHTML★17

  • D AI_Order_to_Cash_Intelligence The name is all big words like AI, BI and Dashboard, with no telling which enterprise process it serves, and no data sources or getting-started notes. Enterprise demo projects like this usually stop at screenshots, and there's nothing verifiable for now. #AI#Data Analysis#Fintech

    AI-powered Order-to-Cash Business Intelligence Dashboard

    AishwariyaecePython★9

  • D elprimerfeca A collection of prompts, layout settings and character settings, a content repo rather than a tool. It only matters to fans of this particular project, with nothing reusable for developers. #Prompting#Content Writing

    Prompts, diseño editorial y biblia de personajes de EL PRIMER FECA.

    SairiAgentJavaScript★23

  • D travel-agent-harness An empty-description Python repo whose name doesn't reveal whether it's a travel agent framework or an evaluation scaffold. With no description it can't enter discussion; skip. #AI Agents#LLMs#Automation

    diva-lPython★17

  • D fsr4xyz The repository name points to something FSR4-related with no description, presumably a way to enable AMD upscaling on unsupported GPUs. It does not say which cards and systems are supported or how to use it, and whether it touches driver or game terms is impossible to judge. #Games#GPU#Performance

    the3rdparty1917Language unknown★11

  • D chai-cli No README body and no description, 14 stars in 1 day, with no telling what it does. The name suggests a CLI tool, but there's no verifiable purpose or usage notes, so treat it as undocumented. #CLI#Productivity

    Aestheticsuraj234TypeScript★14

  • D tdt-nav-kit 45 stars in 0.9 days but no description at all, so it's impossible to tell even which navigation problem it solves, and there's no usage. High stars plus zero explanation looks more like star-farming or a placeholder, so skip it for now. #Robotics#Learning

    T-DT-Algorithm-2026C++★45

  • D moderncpp-course-2627-public A companion materials repo for a modern C++ course with no description. It's course teaching content and assignments, excluded per the rules. #Learning#Build#Code Quality

    cpp-practiceAssembly★12

  • D reel-maker An undescribed TypeScript repository that gained 64 stars in a day with no explanation. Unknown origin; skip it. #Video#Video Editing#Social#AI

    emillycunhaTypeScript★64

  • D snatcher-translated A fan translation patch for the PlayStation version of Snatcher, useful to players wanting this classic text adventure. But the repo has only one sentence plus patch files, with no text extraction, font replacement or patching instructions; the value is all in the patch, and users must supply their own image. #Games#Reverse Engineering

    Fan made Snatcher translation patch for Sony Playstation

    pepasjcLanguage unknown★19

  • D Storm-Dock A Rust project with no repository description; the name suggests some kind of Dock tool, but what it does, who it is for, and how to run it cannot be determined. A likable language choice does not replace documentation, so it cannot be evaluated. #Productivity#UI Design#macOS

    tangsj-hubRust★15

  • D freeclaude A web chat site called "Free Claude" that in substance borrows others' quotas or reverse-engineers APIs to provide access, a way of bypassing payment. It has no technical reference value for developers and carries account and compliance risks; star growth does not make the project itself valuable. #Crack#LLMs#Scams

    free claude cynet ai chatbot

    thecynetxHTML★30

  • D astrable The description cites nonexistent model names like GPT-6 Astra and Claude Fable 5.1, which zeroes out the project's credibility outright. Even if there is a real orchestration idea, there is no way to judge any actual capability from it. #AI#LLMs#AI Agents

    Combine GPT-6 Astra and Claude Fable 5.1. Astra directs, Claude builds, Astra verifies.

    DannyMac180TypeScript★13

  • D student-placement-prediction A typical student assignment: a Streamlit app showing an employment prediction model. Both data and conclusions come from a public teaching dataset, with no reusable engineering value and no discussion of methodological tradeoffs. #Learning#ML Research#Data Analysis

    AI-based student placement prediction using streamlit

    Mavathi11Jupyter Notebook★19

  • D opensas A Zig project whose README has no description, making it unclear what problem it solves. Scored low per the rules. #SDKs#APIs#CLI

    kirha-aiZig★8

  • D levin-harness The repository has no description at all, so its purpose, audience and how to start are unclear. Scored low per the rules. #AI#SDKs#Testing

    levinthalLanguage unknown★12

  • D Perfect-Dark-Plus-HD-Textures An HD texture pack for Perfect Dark, asset distribution rather than a tool, with no description or usage instructions. Possibly useful to players doing emulator HD upgrades, but similar resource sites already offer plenty of alternatives, copyright is murky, and it has no reference value for developers. #Games#Images

    retro-foundryLanguage unknown★20

  • D Debug_time_prediction No description, just a single Notebook that from its name looks like a course or exercise on predicting debugging time. There's no data description, no usable model or interface, and no method worth learning from. #ML Research#Debugging#Learning

    Anubhavsingh311Jupyter Notebook★9

  • D XexTool The repo page has not a single sentence of description, so from the name you can only guess it's an Xbox 360 executable tool. There's no usage and no statement of what problem it solves, and 11 stars provide no evidence; the long-maintained xextool forks already cover this need, so there's no reason to look at a silent repo. #Reverse Engineering#Windows#CLI

    Team-ResurgentC++★11

  • D ai-company-os-github Stacks words like governance, workflow, cognition, orchestration, memory, and security into an 'AI company operating system', with neither an entry point nor evidence. A typical pile of buzzwords. #AI Agents#LLMs#Automation

    portable ai company operating system - executive governance, workflow, workforce, cognition, orchestration, research, memory, quality, security, and autonomous company infrastructure

    karanbinderguptaPython★26

  • D Qwen-Flash-SM80-170HX Claims to be a community vLLM port of a Qwen '3.8-Flash-Next' for dual SM80, but that model name does not exist, and the content is just a string of parameters like GDS, TEP2 and MTP6 with no verifiable build results. No reference value for anyone actually doing inference deployment. #LLMs#GPU

    Community vLLM for Qwen3.8-Flash-Next on dual SM80: GDS, TEP2, MTP6, TileLang, and full-vocabulary draft INT8. Chinese documentation and buildable source.

    nguyenthimy2022kg-altPython★8

  • D Branchglass An HTML project with no repository description, so it is unclear whether it is a tool, a visualization, or a web experiment, and there is no entry-point documentation. There is no information available for judgment. #Web Frontend#UI Design#3D

    wieslawsoltesHTML★9

  • D am-reverse A JavaScript project with no description; the name hints at reverse engineering, but the target platform, analysis subject, and usage cannot be determined. There is no evaluable information, so it can only be skipped. #Reverse Engineering#Debugging#App Security

    zennoptionxctJavaScript★15

  • D minara-feature-videos No description at all; from the name and language it looks like a promo video asset repo for some product. There's neither a tool nor anything reusable, so it has no value for developers. #Video#Marketing#Content Writing

    AntoniaiaiaiaiaHTML★18

  • D MachineLearningCourse2026 A mirror of course materials, with no description. Useful to people taking the course, but not to list readers: it's neither a tool nor a new method. #Learning#ML Research

    MuLabPKUWolfram Language★10

  • D ai-project-learning-skill No description, and from the name it looks like yet another learning-oriented agent skill. There's no way to tell how it's used or what it produces, so it can't be evaluated. #Agent Skills#Learning

    StarYYYzxLanguage unknown★12

  • D gpt5-jailbreak-filter-tester It claims to test jailbreak defenses for a model version that doesn't exist, so the prompt suite itself is unsound. It won't help engineers doing real security evaluation, and it doesn't belong on any list. #Jailbreak#LLMs

    Jailbreak filter tester for AI safety engineers — validate GPT-5.5 Instant Uncensored JailBreak guardrails with repeatable prompt suites.

    SwordPincersHTML★10

  • D neural-code A Python repo with no description, and the name doesn't make clear whether it's neuroscience or a code model. Impossible to evaluate; skip. #ML Research#AI#Science

    avbiswasPython★16

  • D interdimensional-shopping An empty-description JS repo whose name only suggests a small shopping-themed project. With no description there's no basis for judgment; skip. #Web Frontend#Games#Social

    blendi-remadeJavaScript★20

  • D realm-notes An undescribed notes-related JS repo with no explanation of what it solves or who it's for. 20 stars in half a day but zero information; no space. #Documents#Productivity#Web Frontend

    chandanSahoo-csJavaScript★20

  • D sveltekit-books An undescribed SvelteKit repository. Even if the author is a Svelte maintainer, the repo page itself doesn't say whether it's an example, a component, or an experiment, so it can't be recommended. #Web Frontend#Learning#Documents

    dummdidummTypeScript★10

  • D DLSS5-Swapper-Feeder Claims one-click replacement of the DLSS 5 DLL on any GPU and 'DLAA injection', but DLSS 5 does not exist yet, and the description conflates ReShade with neural rendering. Basically a fake tool riding GPU hype, with real risk to players. #Games#GPU#Scams

    Swap DLSS 5 DLLs + Feeder mode for non-DLSS games. One-click DLSS 5 Neural Rendering for any GPU (NVIDIA RTX 20-50, AMD RDNA 3-4, Intel Arc). Synthetic DLAA injection via ReShade.

    roman-g-2098j5Language unknown★17

  • D lunel An empty description, and the repo name conveys nothing either — neither what problem it solves nor how to use it is known. Repos like this have no discussion value on the list. #Learning#Productivity

    ArasTeyPython★11

  • D PVS4_2627 No description, and the name looks like a course or group project number. There's no reusable tool and no method worth reading, so skip it. #Learning#Code Quality

    HonzaBohJava★9

  • D DLSS5-OneClick Another repo from the same source as DLSS5-Universal, also claiming one-click enabling of a nonexistent DLSS 5 on any GPU. Two accounts posting identical content at the same time is typical star-farming; stay away. #Scams#Games#GPU#dlss

    One-click DLSS 5 Neural Rendering for any GPU (NVIDIA RTX 20-50, AMD RDNA 3-4, Intel Arc). Auto-detection, Feeder mode for non-DLSS games, emulator support, and ReShade integration.

    leonmiller20c4Python★19

  • D navin The description piles up AGI, world models, policy learning, autonomous loops and heartbeats, a classic vision checklist. There's no runnable part or concrete usage, making it the kind of project most worth skipping. #AI Agents#LLMs

    Built toward AGI with Persistent Memory, Self-Evolve, Auto-Skills, World Models, Policy Learning, Multi-Agent autonomy, Loops and Heartbeat.

    Navinspire-iaPython★10

  • D scrap-mechanic-menu A game cheat menu repo whose description is a single line of year boilerplate, with no source or technical explanation, typically just funneling to offsite downloads. It falls in the cheat category and is excluded outright. #Game Cheats#Games#Windows

    Scrap Mechanic Mod Menu 2026

    a-clark9057C++★10

  • D Sot-mod-menu A 'premium launcher menu' for Sea of Thieves, i.e., a multiplayer game cheat. The description is vague with no verifiable implementation, so it falls in the excluded category. #Game Cheats#Games#Windows

    Sea of Thieves - Advanced Launcher Menu

    adrianmiller-22Language unknown★12

  • D DLSS5-Universal Claims to enable the nonexistent DLSS 5 on any GPU, with injection mode and emulator support, and a same-named variant appeared around the same time. A typical star-baiting project that may even carry risks—don't download or run it. #Scams#Games#GPU#dlss

    Enable DLSS 5 Neural Rendering on NVIDIA RTX 20-50, AMD RDNA 3-4, and Intel Arc GPUs. One-click setup for any DX11/DX12 game. Auto-detection, Feeder mode for non-DLSS games, emulator support, and ReSh…

    jamesb-io03i8Python★18

  • D Retrac-SDK-14.60 A C++ repo labeled '14.60 sdk and offsets', apparently a set of reverse-engineered offsets for some game version; a gray-area tool. It does not state the game, version or usage, and has no value for normal development. #Game Cheats#Reverse Engineering#SDKs

    retrac's 14.60 sdk and offsets

    pilottX11C++★13

  • D Multichain-Drainer-Tool A multi-chain wallet drainer whose purpose is to move tokens out of other people's wallets, making it a theft tool. We won't discuss the technical details, and it doesn't belong on any list. #Crypto#Scams#Blockchain

    An automated MultiChain wallet drain utility for USDT ETH SOLANA BTC LTC BNB ETC Trc20 trx... tokens bot

    DivideFlyRetakeLanguage unknown★31

  • D pixel-gun-3d-menu Claims to unlock cheats and hidden features in Pixel Gun 3D, with an emoji-stuffed description and no implementation details, offering zero value to developers and violating the game's terms. A typical cheat-traffic repo. #Game Cheats#Games#Android

    🕹️ Unlock cheats and strategies for Pixel Gun 3D on PC to enhance your gameplay and discover hidden features easily.

    a-henderson18Language unknown★10

  • D arma-reforger-mod-menu Claims to use 'advanced techniques and reverse-engineering experience' for an Arma Reforger mod menu, but it's actually a multiplayer cheat tool. The copy is inflated, no source is disclosed, and nothing is verifiable, so it's excluded outright. #Game Cheats#Games#Windows

    Truly unique Mod Menu for Arma Reforger developed using advanced technologies and reverse expertise. We proudly present a reliable solution that will give you an edge over your competitors.

    abigails2572Language unknown★10

  • D core-keeper-mod-menu Just one line, 'Core Keeper Mod 2026,' a cheat tool repo with no description or implementation details. A typical seasonal star-farming placeholder project; no score. #Game Cheats#Games#Windows

    Core Keeper Mod 2026

    adammorales325Language unknown★10

  • D rainbow-six-menu An ESP and no-recoil suite for Rainbow Six Siege with a description that's all keyword stuffing. A cheat, excluded outright. #Game Cheats#Games#Windows

    Top Rainbow Six Siege Gameplay Enhancement Suite 2026: ESP & No Recoil

    adriford1022Language unknown★10

  • D mythos-5-incident-transcript An HTML content repo under the anthropics org name titled 'incident records,' with no description. Authenticity and provenance can't be verified, and it looks like a name-borrowing fake content page, so it's not considered. #Scams#LLMs#Documents

    anthropicsHTML★11

  • D how-to-fish-BSProject-ModMenu Titled 'tutorial' but actually a repost of some game's ModMenu, with the description admitting it's someone else's modifications and additions. It's cheat-category content, excluded. #Game Cheats#Games#Windows

    How to Fish "BSProject - ModMenu by QuasarUI & modified|added by blindextent

    blindextentLanguage unknown★10

  • D TRX-Drainer-Tool This is a TRON wallet drainer that automatically extracts TRX/TRC20, i.e. malware. Its 39 stars in 1.0 days may be gray-market hype, with no legitimate commercial value; legal risk is extremely high and it harms users' assets, so it is absolutely excluded from watch. #Scams#Privacy

    TRX Drainer Tool — Automated TRX and TRC20 token drainer for TRON wallets with real-time balance monitoring, multi-wallet sweeping, TronLink/Trust Wallet compatibility, and stealth energy-optimized tr…

    joanmach77Python★39

  • D geometry-dash-mod-menu A Geode mod menu for Geometry Dash whose entire description is a pile of geode keywords and platform names; a typical keyword-farm mod/cheat repo. No readable technical content for players or mod authors. #Game Cheats#Games#Scams

    Geode GD Mod Menu geode geometry dash, geometry dash geode, gd geode, geode gd, geode download, geode installer, how to install geode, geode 2.208, geode apk, geode mods, geode github, geode sdk. Wind…

    oevans8933Language unknown★14

  • D jailbreak-script-2026-heist-route-toolkit A so-called script toolkit named after Roblox's Jailbreak, bundling 'robbery route planning', 'vehicle tracking' and scripts together; essentially a cheat/traffic repo riding game keywords. No value to normal players and against platform rules. #Game Cheats#Games#Scams

    Roblox Jailbreak Script 2026 themed toolkit for heist planning, vehicle tracking, cash progression, routes, loadouts and gameplay analytics.

    ncollins3730Language unknown★10

  • D stalker-2-trainer-cheats-mod-toolkit A game cheat mod pack under the name of an S.T.A.L.K.E.R. 2 trainer, dressed up with gameplay words like exploration and weapon builds but actually pointing to cheats. No legitimate use for players and against game terms; a keyword repo not worth the time. #Game Cheats#Games

    S.T.A.L.K.E.R. 2 Trainer themed PC toolkit for Zone exploration, weapon builds, artifacts, inventory management, progression, mods and gameplay analytics.

    noah2001-artLanguage unknown★10

  • D Hunt-Showdown-1896-Ultimate-Assistant A cheat tool for Hunt: Showdown 1896 explicitly listing 25 features such as ESP, aimbot and God Mode. It ruins other players' matches and violates game terms, with nothing worth a developer's time. #Game Cheats#Games#Anti-Cheat

    Ultimate assistant for Hunt: Showdown 1896. Features: ESP, Aimbot, Radar, God Mode, Unlimited Health/Stamina/Ammo, Super Speed, and more. 25+ functions.

    samm-io2094x8Python★9

  • D Steal-an-Egg-Roblox The repo name points to Roblox's Steal an Egg, with no description or instructions; it is impossible to tell whether it is a tool or a script injector, or what it is for, so it can only be skipped. #Games

    solodisaLanguage unknown★14

  • D instagram-reporting-bot A bot for mass-reporting Instagram accounts, effectively weaponizing the platform's reporting mechanism. There's no technical novelty, and it shouldn't be recommended given its purpose. #Scams#Social

    Bots are automated tools designed to streamline the process of submitting reports on Instagram accounts that may violate the platform’s Community Guidelines. They can assist with reporting issues such…

    SolicitorDeemRuby★31

  • D facebook-reporting-bot A bot for mass-reporting Facebook accounts, essentially an attack tool that abuses the platform's reporting mechanism. There's no technical novelty, and it shouldn't be recommended given its purpose. #Scams#Social

    Bots are automated tools designed to streamline the process of submitting reports on FaceBook accounts that may violate the platform’s Community Guidelines. They can assist with reporting issues such …

    EmblemPatronBorderLanguage unknown★12

  • D ChatGPT-i18n An empty-description repo whose name hints at ChatGPT interface internationalization, with no implementation or usage notes. The same account shipped multiple repos in one day with highly similar star counts, an obvious star-farming pattern, so all are excluded. #LLMs#AI#Web Frontend

    ashelygierischTypeScript★60

  • D ESP32-AI One of a batch of repos from the same account, tagged ESP32 and AI but with no hardware model, model details, or run instructions. Real ESP32 inference projects come with clear docs and examples; this has only a name and star count. #Embedded#Hardware#AI#IoT

    ashelygierischPython★62

  • D React_dash_tremor_Tailwind The repo name is just a stack of technologies with an empty description, a template or practice project. Multiple repos from the same account got similar star counts on the same day, a clear star-farming signature that isn't a signal worth looking at. #Web Frontend#Data Analysis#UI Design

    ashelygierischJavaScript★66

  • D Time_plottimeseries_CLI_tool A time-series plotting CLI with only a repo name, no description or examples. The same account's batch repo creation and star-farming pattern is obvious, and there's no way to tell if it runs, so skip it. #CLI#Data Analysis#Terminal

    ashelygierischTypeScript★62

  • D chatify A chat interface practice project with an empty description. It and the account's other repos gained roughly 60 stars within half a day to two days, highly suspicious, so it isn't taken seriously. #Social#Web Frontend#AI

    ashelygierischJavaScript★60

Show HN 5

S

Show HN: Geiger – See every AI agent on your machine and what it can touch

Show HN

Addresses the difficulty of tracking proliferating AI agents with a read-only command that inventories agents, MCP servers and plugins. Its 44 HN points and 68 stars show strong security-community interest; commercialization means part of a security audit or asset management tool or a SaaS offering, but the AI ecosystem changes fast and operating systems may build similar features.

#AI Agents#App Security#Privacy#MCP

▲44

S

Show HN: Self-hosted company OS, Claude Code and Codex agents in departments

Show HN

A self-hosted multi-tenant AI agent platform where teams collaborate with Claude Code/Codex inside sandboxes. It gained 95 stars in 0.3 days and 37 HN points, strong momentum; commercialization resembles a company OS with team subscriptions or private deployment, but competition is fierce and dependence on Anthropic/OpenAI subscriptions exposes it to platform policy.

Hi HN, I am Dimitris. This is a company OS that I built and use to run my business and anyone can install it and self host it for free. Think of it as Claude Code, Cowork and the cloud sessions in one…

#AI Agents#Self-Hosted#Collaboration#claude-code

▲37

S

Show HN: Give your AI agent on-screen guides that show users where to click

Show HN

Solves AI agents not understanding their own product UI by using a browser agent to learn app structure and generate on-screen guidance. Its 17 HN points show interest, from a YC startup; commercialization means selling an assist API as SaaS billed per call, but RAG keeps improving, enterprises may self-train, and browser-agent security and maintenance costs matter.

Hey HN. I'm Christian, one of the founders of Frigade (YC W23). I've noticed that a lot of in-app AI agents struggle to actually understand the products they exist in. For instance, let's say a user a…

#AI Agents#APIs#SDKs#UI Design

▲17

A

Show HN: Ctrlb-decompose: Strip the noise from logs before sending to LLMs

Show HN

Solves logs drowning LLM analysis in noise by decomposing logs into a surface usable for reasoning. It gained 297 stars in 0.5 days, hitting an AI debugging pain point; commercialization means integration into observability or AI debugging tools or becoming an open-source component standard, but existing log platforms may build similar features or the project may stay too niche.

#LLMs#Debugging#Observability#Monitoring

▲13

A

Show HN: Type.com: Multiplayer Codex/Claude in the cloud for non-tech use cases

Show HN

A multiplayer AI agent cloud IDE for non-technical users, built on Claude Code/Codex and usable in Slack. Its 14 HN points come from a very early stage, but positioning is clear and commercialization prospects are good; risks are direct competition with Claude Cowork and costly VM infrastructure, and the path to a standard tool is team collaboration stickiness.

Hey HN, I'm Komran, the CTO and cofounder of type.com. We're launching today. I've been describing type as "the IDE for non technical people." It allows you to build really cool shit without having to…

#AI Agents#Collaboration#Productivity#claude-code

▲14

Hot New Repos

  1. openai/NavierStokesAndEuler★1551 Lean certificates accompanying Navier-Stokes and Euler results
  2. gazijarin/itsgiving★481 Express yourself in meetings (with memes, of course).
  3. com-phantom/Phantom-Wallet-Desktop★451 Phantom Wallet Desktop app for Windows, macOS and Linux. A lightweight desktop client that brings Phantom crypto wallet experience outside the browser with native desktop access.
  4. henryzawadzki6542/cloudflare-turnstile-bypass★358 Work with Cloudflare Turnstile in Python: find the sitekey, get a valid cf-turnstile-response token in seconds. Dependency-free library + CLI, built for CI and automation.
  5. nagisanzenin/truanayangi★301 CS:GO-style lunch roulette. Mở hòm, chốt món ăn trưa.
  6. tristanbuckmaster/fluid_lean★233
  7. jaskirat1616/mactap-app★230 MacTap — knock your MacBook to run shortcuts
  8. jnz/INSLIB★188 Open Source Inertial Navigation Library
  9. Foadsf/vintage-latex★125 Twenty LuaLaTeX and MetaPost (fiziko) examples of the old scientific-paper aesthetic: engraved plates, star charts, computed log tables, knots, telescopes
  10. andrewroxby/claude-style-patch★122 A drop-in CLAUDE.md style section that changes how Claude writes prose.

Star gain 1d

  1. openai/NavierStokesAndEuler+484★1551
  2. sdli1995/dlssg_for_sm86+438★657
  3. larashero3-dotcom/lieflat-gongwen+373★546
  4. com-phantom/Phantom-Wallet-Desktop+328★451
  5. donvito/codex-astra-luna-orchestrator+296★717
  6. ashemag/human-atlas+295★2836
  7. XiaoDuoYa/codex-with-chatgpt+223★3768
  8. lnkiai/m3e-canvas+217★5530
  9. rakanki911/DLSS5-Swapper+191★3256
  10. ApodexAI/FrontierAgent+174★2435