tierlist.tech
中EN

10.09 Today's Picks

349 new projects found today, 189 rated.

Created 2026-10-07 – 2026-10-09Fetched 600Filtered 50Rated 26

S · Look today

S

MacMiniMode

Pama-LeeSwift

Keeps a MacBook awake with the lid closed and no external display, auto-disabling on battery — exactly the gap in macOS. One sentence states the use and limits, it's native Swift, and 100+ stars in two days shows real demand; the limit is that it only serves MacBook users.

Keep a plugged-in MacBook awake with the lid closed: clamshell mode without an external display, auto-off on battery. Let a MacBook stay awake with the lid closed and use it like a Mac mini.

#macOS#Productivity#Automation

★119

A · Worth watching

A

ARTEX

mhtsecGo

An autonomous AI penetration testing system and Baidu agent+ attack-defense competition champion, with 696 stars in 0.8 days, the strongest momentum in this batch. It chains reconnaissance, exploitation, and reporting into a closed loop, directly targeting pentest-as-a-service and enterprise red-team subscriptions with the clearest commercial path. Risks are liability for accidental damage and pressure from big-vendor security platform bundling.

AI autonomous penetration testing system | Champion project of Baidu's "agent+" attack-defense challenge

#AI Agents#Network Security#Exploits

★696

A

answerui

0xcro3dileTypeScript

A generative UI layer that renders model replies as charts, forms and calculators, compatible with any OpenAI-compatible model, suited to indie developers building interactive answers. Lighter than frameworks like CopilotKit, but built in one day, so component reliability and fallbacks for unstable model output need real testing.

Open-source alternative to ChatGPT's Intelligent UI. Answers arrive as live charts, forms and calculators. Works with any OpenAI-compatible model, in the cloud or local.

#LLMs#UI Design#Open-Source Alternatives#Web Frontend#chatgpt

★23

A

WinShun

LingCoreC++

A Windows quick launcher summoned by double-tapping Ctrl, supporting filename, pinyin and full-text search in Qt6 + C++20, with clearly stated scope. Everything and Listary are rivals; its edge is Chinese pinyin and content search, and large-index responsiveness needs real testing.

A fast search launcher for Windows: double-tap Ctrl to summon, search by filename, pinyin and text content. Qt 6 + C++20.

#Windows#Productivity#Automation

★38

A

agentic-kb

beianchuqing111Python

A knowledge-base agent for regulatory documents that chains dense and sparse retrieval, RRF, cross-encoder reranking, GraphRAG and audited ReAct, plus retrieval evaluation tooling. Teams building enterprise compliance Q&A can borrow from it, though generality depends on the corpus.

Regulation/procedure knowledge base agent: hybrid retrieval (bge-m3 dense + sparse → server-side RRF → cross-encoder reranking) + GraphRAG + a ReAct agent with three gates and auditing. Includes retri…

#RAG#AI Agents#Knowledge Base#LLMs

★58

A

cli

excalidrawTypeScript

Excalidraw's official CLI that script-generates and processes .excalidraw diagrams, also serving agent workflows. Official and restrained in scope; the only issue is it's one day old and interfaces will still change.

Official Excalidraw CLI for humans and agents

#CLI#Editors#AI Agents#excalidraw

★31

A

PiDurableKit

finnvoorSwift

It brings durable execution to Apple platforms, running agent logic on a JS engine with Swift-side APIs—same lineage as Cloudflare Workflows but on-device. The scope is clear and momentum fast; worth seeing how it handles checkpointing and recovery.

Durable AI agents for Apple platforms: pi-durable on JavaScriptCore with a Swift API

#AI Agents#SDKs#iOS#macOS#javascriptcore

★70

A

betterwispr

opennookorgSwift

Local dictation for macOS: hold a hotkey, speak, and it types, supporting Whisper, Parakeet and Apple speech, explicitly targeting Wispr Flow and fully offline; the need is clear and onboarding straightforward, so it comes down to whether recognition quality and latency can match the commercial version.

Free, open-source, local-first dictation app for macOS. Hold a shortcut, speak, and text is typed into any app. On-device speech-to-text with Whisper, Parakeet and Apple speech. Private, offline, open…

#Speech#macOS#Open-Source Alternatives#Privacy#wispr-flow

★49

A

CF-Workers-TGProxy

ToiCFLanguage unknown

It runs a Telegram Web/MTProto WebSocket relay on Cloudflare Workers, with bridge auth, multi-lane multiplexing, AES-CTR and flow control implemented, so it's not an empty shell. Useful for users behind network restrictions and those learning Workers long connections; deployment and compliance are up to you.

A Telegram Web/MTProto WebSocket proxy relay based on Cloudflare Workers, supporting bridge auth, bootstrap/session tokens, multi-lane WS multiplexing, AES-CTR encryption, custom frames, flow control …

#Networking#Privacy#Network Security#telegram#cloudflare-workers

★46

A

map-motion

SpaceZephyrJavaScript

A Claude Code skill that turns real Amap routes into map-animation videos, with 12 styles including dive, altitude and flight path, and a clear positioning. Travel or cycling vloggers can produce clips fast; the limit is being locked to Amap data and the Claude Code ecosystem.

Map motion-graphics video Claude Code skill: based on real Amap routes/boundaries/satellite basemaps, creates map animations such as Earth dive, vlog "I am here", hiking elevation, cycling journey, fl…

#Agent Skills#Video#Animation#amap

★15

A

PaperBenchX

UniPat-AIPython

A benchmark evaluating whether research agents can reproduce papers end to end, covering 12 research fields, hitting the gap in verification for AI-driven science. Worth a look for those doing agent evaluation or research automation; the limits are that question sourcing and scoring criteria need close reading.

A benchmark for scientific research agents, evaluating end-to-end paper reproduction across 12 research areas.

#Benchmarks#ML Research#AI Agents#Science

★19

A

freecenter

fdafornoTypeScript

It adds a web admin console for free ESXi, which lost vCenter after Broadcom, with migration, backup and a Terraform provider. The pain point is real, but claiming so much in just over a day—check the code before trusting it.

A vCenter-style management console for the free VMware ESXi hypervisor: web UI, REST API, migration between hosts, backups, Terraform provider

#Self-Hosted#DevOps#APIs#vmware#terraform

★28

A

streamgres

juspayRust

Juspay's Postgres subscription-based sync engine, built on logical replication for cross-database sync. The company backing and 66 stars suggest real demand, but how it beats PeerDB-style options requires reading the code.

A subscription-native sync engine for PostgreSQL.

#Databases#DevOps#postgres

★66

A

tailscale-nx

UnderwaterOvergroundRust

It turns Tailscale into an Atmosphère system module so homebrew like Moonlight on the Switch can join the tailnet directly, a concrete idea with clear boundaries. People streaming on hacked consoles will want it immediately; the limit is serving only already-jailbroken Switches.

Tailscale for the Nintendo Switch: an Atmosphère sysmodule that puts homebrew (like Moonlight) on your tailnet

#Networking#Games#tailscale

★15

A

It decouples the parameters of context writing and denoising in autoregressive video generation, claiming a 5-second training window can generate up to 24 hours. The idea is concrete with complete code, an extension of the Self Forcing line. Whether it holds depends on VRAM and quality costs; suitable for long-video generation people to reproduce.

Autoregressive video generation with decoupled context-writing and denoising parameters, enabling up to 24-hour generation from 5s training windows.

#Video#ML Research#AI

★55

A

tf2mt

HorrorPillsC

Runs TF2 on Apple Silicon via Wine+DXVK+MoltenVK, listing tested fixes (mouse feel, menu stutter) plus a native launcher — directly useful to Mac players. The limit is that it serves one old game and isn't reusable across titles.

Team Fortress 2 tuned for Apple Silicon: Wine + DXVK + MoltenVK with measured fixes (smooth 120 Hz mouse, no menu stalls) and a native macOS launcher

#macOS#Games#Game Porting#wine#tf2

★11

A

ApiHashing

Maldev-AcademyC

A Windows x64 API hashing library that bypasses EAF and resolves forwarded exports, ordinals and api-ms-* API Sets, from Maldev Academy — practical for red teaming and malware development. The description is precise and unhyped, but the audience is very narrow and it's a dual-use tool.

Windows x64 API hashing library that bypasses EAF and resolves forwarded exports, ordinals, and api-ms-*/ext-ms-* API Set DLLs.

#Windows#Reverse Engineering#App Security

★34

A

stocking-texture-tool

silvermoongPython

Adds moiré-free knit texture to stockings in illustrations, a narrow but real problem, described accurately in both Chinese and English without hype. People doing AI image post-processing will want to try it; the limitation is it covers only this one material, so generality is limited.

Adds a realistic, moire-free knit texture to the stockings in an illustration.

#Image Generation#Images#AI

★28

A

MirrorAct

sopitzSwift

Mirrors iPhone/iPad screens to a Mac with device frames plus screenshots and recording, a real need for developer demos and App promo material. QuickTime only works over USB, so the AirPlay part is the differentiator; 11 stars is too early and stability remains to be verified.

Mirror, frame and record your iPhone or iPad on your Mac – via USB or AirPlay. Device frames, backgrounds, screenshots, recordings, presentation mode. Open source (GPL-3.0).

#iOS#macOS#Video

★11

A

OpenComputerUse

IAmJSDRust

An MCP server that gives agents background computer control, driving the desktop without stealing the foreground, written in Rust with a narrow scope. The pain point is real, but it's one day old, so the permission model and long-task stability are unverified.

Background computer use for agents, as an MCP server

#AI Agents#MCP#Automation

★10

A

loadout

MK27MKTypeScript

A harness manager for Claude Code configs that switches by profile, shows where each setting comes from and toggles them individually, solving the real pain of config sprawl. Scope is restrained, but it depends heavily on Claude Code's config format and must chase upstream changes.

Claude Code harness manager: switchable profiles, scope provenance and on/off for every harness element

#AI Agents#CLI#Productivity#claude-code

★13

A

MathModel-Copilot

OdyphusPython

A math-modeling assistant and agent skill for CUMCM and MCM/ICM, covering problem reading, model comparison, running Python, result verification and paper support, matching real contest rhythm. The audience is competing students and the scope is clear; similar tools abound, so the key question is whether result verification is truly reliable.

A mathematical modeling AI assistant and Agent Skill for CUMCM and MCM/ICM: problem analysis, model comparison, Python execution, result verification, paper assistance and team collaboration. Mathemat…

#AI Agents#Agent Skills#Learning#LLMs

★40

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

BHas a spark44
  • B fsearch Full-disk fuzzy search for macOS, scanning 8 million files in about 1ms with content grep, directly filling gaps where Spotlight is slow and Raycast coverage is incomplete. With 389 stars in 0.3 days, momentum is solid and developers pay willingly, so it could follow the Mac one-time-purchase path. The risk is Apple improving indexing APIs at the system level at any time. #Productivity#macOS#Performance

    Whole-disk file search for macOS: fuzzy names, typo tolerance, indexed content grep. ~1 ms over 8M files.

    noahdunnaganRust★389

  • B cirfig_gen Lets Claude Code or Codex generate final-size analog-circuit SVGs on demand — a very narrow but clear niche for people writing papers and circuit docs. The risk is whether generated symbols are standard and dimensions reliable, which needs testing on real circuits. #AI#Images#Hardware#CLI

    AI-driven analog circuit figure generator. Ask Claude Code or Codex for a circuit and get one SVG at final size.

    KwantaeKimPython★11

  • B wireztna-oss Self-hosted zero-trust access built on WireGuard, authorizing per user rather than per device and requiring no domain, filling a gap left by Tailscale-style offerings. The description is restrained and clear; with 11 stars in 0.6 days, whether policy distribution and management are complete remains unverified. #Network Security#Networking#Self-Hosted#Privacy

    Open Source, Self-hosted Zero Trust Network Access (ZTNA) built on WireGuard. Fine-grained per-user access to private resources — no domain required.

    wireztnaGo★11

  • B HookRelay A self-hosted webhook receiving and multi-target forwarding gateway with HMAC verification, persistent queues, retries and dead letters, aimed at small teams needing reliable delivery. The scope is restrained and the description concrete without hype, but mature alternatives like Convoy exist, so documentation depth decides whether it's truly usable. #APIs#DevOps#Self-Hosted#Network Security

    Security-hardened webhook ingestion & multi-destination delivery gateway (GitHub ➔ Discord, Slack, HTTP) with HMAC-SHA256 verification, durable SQL queueing, fencing leases, smart retries, and DLQ.

    Abhishek-GaliPython★10

  • B pluck Wraps yt-dlp in a native macOS interface with Spotify playlist support, friendly for non-CLI users. But many such GUI shells exist, the Spotify part likely only fetches metadata, and differentiation is limited unless its update and concurrency handling is genuinely more stable. #macOS#Downloaders#Video#yt-dlp#spotify

    A native macOS app for yt-dlp, with Spotify support and self-updating tools

    AstrofroggerSwift★29

  • B welp A small tool that stops you before you send a message to the wrong window, with an accurate scenario and restrained tone that will resonate with people juggling many chat windows. But it doesn't say which clients it supports or how it detects the wrong person, and 17 stars isn't enough to judge, so note it for now. #macOS#Productivity#Privacy

    Your boss didn’t need to see that. Welp stops you before you send to the wrong chat.

    actuallyzefeSwift★17

  • B visionflow-studio A low-code machine-vision platform for industrial inspection, with 110+ operators, PLC/Modbus communication and export to HDevelop and C# code, clearly built from real production-line projects. Engineers in industrial vision should evaluate it, but platform projects risk piling on features, so check operator depth first. #Images#Automation#IoT

    visionflow-studio: a low-code machine vision platform for industrial inspection. Drag and drop 110+ vision operators to build pipelines, AI assistant helps tune parameters, supports PLC and Modbus com…

    agentthinkC#★18

  • B simpsons-wrestling-recompiled It uses psxrecomp to statically recompile PS1's The Simpsons Wrestling to Windows and Linux, with widescreen, 60fps and unlocked content, and honestly requires your own disc. Worth reading for those interested in static recompilation pipelines; the limit is a single game and a nostalgia-oriented audience. #Reverse Engineering#Game Porting#Games

    The Simpsons Wrestling (PS1, USA) recompiled for Windows and Linux with psxrecomp and the TRG Launcher: 16:9, HD, 60 FPS, skip intro, everything unlocked. Bring your own disc.

    TekRantGamingC++★14

  • B nx-on-orbis It ports the Switch emulator Eden to a jailbroken PS4 Pro, openly stating it's a discontinued experiment and providing no keys or games, a rare honesty. The technical approach is a reference for emulator developers, but the repo itself isn't directly usable. #Reverse Engineering#Game Porting#Games

    Experimental port of the Eden Switch emulator to a jailbroken PS4 Pro (closed experiment, open for anyone to continue). Not affiliated with Eden; built with AI. No keys, firmware or games.

    alechurriC++★16

  • B Pawprint A client-side build planner for Minecraft with ghost blocks, placement previews, and a blueprint library, with concrete features listed. Litematica is already the de facto standard here, so it must win on lighter interaction; 94 stars in 1.7 days is real heat but needs verification. #Games#Game Dev#Productivity

    Client-side building planner for Minecraft: ghost-block blueprints, placement overlay, studio world and blueprint library

    yujin2625Java★94

  • B UltraText_Bench A bilingual benchmark for text rendering in image-generation models, filling a real gap with concretely defined scoring dimensions. But it's a lab evaluation artifact that ordinary developers can't use today — better suited to image-generation researchers. #Benchmarks#Image Generation#ML Research#Multimodal

    UltraText Bench: A Comprehensive Bilingual Benchmark for Evaluating Visual Text Rendering in Image Generation

    LINs-labPython★15

  • B Motion-graphics-skill Motion-graphics and video-editing skills for Claude Code, Codex and Cursor, claiming local rendering as an After Effects replacement. It calls itself 'Best' with an oversized capability list but no sample output or rendering-quality evidence — currently more of a marketing page. #Agent Skills#Video Editing#Animation#LLMs

    Best Motion graphics and video editing skill for Claude Code, Codex and Cursor. Plan, animate, edit, caption and sound-design complete videos, rendered locally without After Effects.

    Miftahul-Islam-EfazShell★25

  • B YANchor It claims a 4B model does long-range reasoning in O(N) time and O(1) memory, and ships inference and evaluation code. The direction is worth a look, but such complexity claims usually depend on specific settings and no comparable baseline is given, so long-context people should verify it themselves. #LLMs#ML Research#Performance

    YANchor-4B: Effective Long-Horizon Reasoning in O(N) Time with O(1) Memory. Inference and evaluation code.

    RocoreMatrixPython★20

  • B manufacturing-forecasting-framework A reproducible pipeline for monthly manufacturing demand forecasting: random forest, time-based backtesting, a naive baseline and a small local app, with decent engineering judgment. But such templates are common, and seventy-plus stars in a day doesn't match the niche topic; read the code before trusting the stars. #Data Analysis#Automation#Benchmarks

    Reproducible monthly demand forecasting with Random Forest, chronological backtests, transparent baselines, and a guided local app.

    Yu-ChangChengPython★73

  • B multi-database-mcp-server A read-only MySQL/PostgreSQL MCP server in TypeScript + TypeORM + Docker that plugs straight into Claude and similar clients, with clear boundaries. But several official and community equivalents exist and it states no differentiator, so the longer-maintained ones are the safer pick. #MCP#Databases#DevOps

    Read-only Model Context Protocol server for MySQL and PostgreSQL, built with TypeScript, TypeORM and Docker.

    HaticeGuzellTypeScript★14

  • B drama-series-agent It strings a one-line prompt through short-drama scripting and casting to MiniMax rendering in one pipeline, a clear product idea riding the AI short-drama trend. But most one-line-to-video projects stop at demos, so output quality and reusability are questionable. #AI#Video#AI Agents#LLMs

    A one-line-to-video short-drama series Agent (Series-scoped agent: idea → short-drama script → cast → MiniMax H3 Ref2VA render)

    binwu1Python★72

  • B uni-switch A cross-platform switcher for Codex/Claude API configurations under AGPL, removing the hassle of editing multiple keys and endpoints. The function is clear, but many similar small tools appeared this month, differing mainly in UI. #LLMs#Productivity#CLI

    Open-source Codex and Claude API configuration manager for Windows, Linux and macOS. AGPL-3.0-only.

    dieqiyunRust★22

  • B windows-partition-manager An open-source tool that carves partitions from free space using only the Windows Storage Management API, cleanly scoped to fill the gap beyond commercial partition software. First confirm it safely handles existing partition tables. #Windows#Productivity#Hardware

    Free, open-source Windows partition manager: carve new partitions out of free space safely, using only the Windows Storage Management API. C# / .NET 10 / WPF.

    kylehesiC#★10

  • B weasley-clock-card A Home Assistant card mimicking the Weasley clock, with one hand per person showing their location — a concrete idea with clear scope, a clean little tool. But it only helps HA users who already have family location entities set up, so the audience is too narrow. #UI Design#Web Frontend#home-assistant

    Home Assistant card in the style of Mrs. Weasley's clock: one hand per person, pointing at where they are.

    Nite01007JavaScript★14

  • B zig-secretstack A small library by jedisct1 that keeps keys out of memory during sensitive operations, with a one-line purpose. The author is credible and scope restrained, but the description is too short; at this stage it's more an implementation for peers than a usable tool. #Secrets#Privacy#App Security

    Run sensitive operations without leaving secrets around.

    jedisct1Zig★11

  • B mxu-slides A native macOS presentation app for churches — clear scenario, small scope, no hype in the description. But the audience is very narrow and mature commercial options like ProPresenter already serve it, with no visible reason to switch. #macOS#Presentations#UI Design

    MxU Slides: a native macOS presentation app for churches

    MxU-EngineeringSwift★11

  • B rapido-tts A 21M-parameter Brazilian Portuguese TTS focused on being small and fast, with a clear positioning and no hype. Developers working on pt-BR speech can use it as a lightweight option; the limits are single-language coverage and no comparison data against other small models. #Speech#AI#Performance

    Tiny, fast Brazilian Portuguese text-to-speech (21M params)

    anurag12-websterPython★9

  • B prism-viz It runs sandboxed, theme-following interactive charts inside Obsidian notes, with isolation and coloring considered—a clean small plugin. The audience is limited to Obsidian users, and at 10 stars its rendering limits are unclear. #Plugins#Editors#Data Analysis#obsidian

    Prism: sandboxed, themed, interactive visualizations in Obsidian notes

    floorianmbTypeScript★10

  • B scooby-doo-mistery-mouse-version It reverse-engineers logic from Genesis cartridge 68000 code and rewrites it in SFML as a mouse-playable native port, requiring your own ROM. Technically skilled and well-judged, with a very narrow audience—worth ten minutes to see how it's dissected. #Game Porting#Reverse Engineering#Games#sfml

    A native SFML port of the Sega Genesis point-and-click adventure Scooby-Doo Mystery, reverse-engineered from the cartridge's 68000 code and playable with a mouse, that runs from your own ROM.

    janezkranjcC++★12

  • B docflare-ai Adds cited Q&A to documentation sites with a single script tag, running full-stack on Cloudflare's free tier; the tradeoffs are pragmatic, but 'AI search for docs' is already a red ocean, leaving deployment cost as the only differentiator. #RAG#LLMs#AI#Web Frontend#cloudflare

    Open-source AI chatbot & site search for your docs. One <script> tag, streaming answers with sources. 100% Cloudflare-native (Workers AI + Vectorize + D1), runs free on the free tier.

    p10nodeTypeScript★24

  • B Nested-Adversarial-Containment An experimental framework for studying multi-agent breakout and isolation under survival pressure, sitting squarely on the hot topic of agent safety. But it only describes the framework, not how to run it or what results it produced, so for now it reads like paper-accompanying experiment code. Worth a look for those following multi-agent safety. #AI Agents#ML Research#App Security

    Empirical evaluation framework for multi-agent safety, adversarial breakout, and containment dynamics between autonomous LLMs under operational survival pressure.

    PranilShah-08Python★10

  • B multi-codex-app It lets multiple Codex desktop accounts log in in parallel with separate profiles, solving multi-instance needs under rate limits. The implementation is direct, but multiple accounts may violate terms of service, and its value depends entirely on someone else's product. #LLMs#AI Agents#Productivity

    Run several Codex desktop accounts side by side, each with its own separate profile and login.

    am-willGo★20

  • B OpenSLA A release of an open sensor-language-action model; VLA is exactly where robotics most lacks open weights. But the description is one line, with no model size, data, or license, so researchers should glance at it while others cannot use it yet. #Robotics#ML Research#AI

    Open Sensor-Language-Action (SLA) Models

    yang-ai-labPython★17

  • B drifting-tts One-step Turkish TTS using Kyutai's learned-temperature recipe, with language and inference steps as gaps. It's research-flavored with little ready to use—good for studying the method rather than as a product. #Speech#ML Research#AI#kyutai

    One-step Turkish text-to-speech trained with a drifting objective (Kyutai learned-temperature recipe)

    kadirnarPython★17

  • B integer-mult-kappa It gives a concrete conditional witness constant for integer multiplication, continuing related discussion by OpenAI and Colkitt, stated precisely and very narrowly scoped. Only those interested in verifiable computation and complexity lower bounds will find it useful, and the conclusion has not been independently checked. #Crypto#Science#Benchmarks

    Conditional witness kappa = 249/(5*10^10) for integer multiplication, building on OpenAI #109 and Colkitt's integer-mult-bounds

    Swapnil-jainPython★14

  • B AiOperationsRoom A local panel for watching multiple Claude Code sessions and sub-agents at once, solving real multi-session chaos; but similar monitoring panels have clustered recently, and the Portuguese description doesn't explain how to connect. #AI Agents#Monitoring#Observability#claude

    Sala de operações local para acompanhar sessões e subagentes do Claude Code.

    mewsdevJavaScript★30

  • B tayga Turns OpenTelemetry traces and logs into a readable real-time incident narrative, a more interesting idea than yet another dashboard. But there's only one sentence of description, with no clarity on output form or integration cost, and 14 stars in 1.3 days shows no real users yet. #Observability#Debugging#DevOps

    Real-time error stories from OpenTelemetry traces and logs (Rust)

    softberriesRust★14

  • B ShutterSoundToggle Uses Shizuku to disable the forced shutter sound, a real problem on Japanese and Korean models, implemented with restraint. But it removes exactly the protection local legislation created against covert photography, so it's hard to recommend. #Android#Privacy#Automation

    Toggle the enforced camera shutter sound through Shizuku

    soralis0912Kotlin★9

  • B pocketvibe A three.js game launcher and store for retro handhelds, plus an AI starter template for making your own games—an interesting direction with a clear audience. But the ecosystem is early, and stable running on real ROCKNIX hardware needs verification. #Games#Game Dev#Web Frontend

    Play three.js games on your retro handheld, and make your own with AI. A launcher, game store and starter kit for ROCKNIX and Android handhelds.

    cobanovJavaScript★9

  • B winkit A Windows 10/11 tuning script collection emphasizing that every change is reversible, which is more honest than peers. But Chris Titus's WinUtil already owns this space, and rollback alone rarely convinces people to switch. #Windows#Productivity#Automation

    Open-source toolkit for setting up and tuning Windows 10 and 11. Every change can be undone.

    klaudikdevPowerShell★27

  • B mcpdelta Aims to add a layer between AI apps and MCP servers, combining multi-step tool calls into one program to save tokens with rollback logging; the idea is interesting, but all three platforms are 'coming soon' so it can't be installed now, and the 24.1x figure is just marketing for now. #MCP#AI Agents#LLMs#Benchmarks

    Delta MCP is a free app that sits between your AI apps and their MCP servers: one program per task instead of one tool call per step, up to 24.1× fewer tokens in our benchmarks. Every change is record…

    mcpdeltaLanguage unknown★79

  • B nodecloak Creates isolated browser profiles for Windows/macOS with a separate proxy per profile, and also launches Claude and Codex terminals; the feature description is specific, but fingerprint-isolation tools are often used for multi-account operations, leaving positioning and compliance boundaries vague. #Privacy#Windows#macOS#Networking

    NodeCloak for Windows and macOS: isolated browser profiles, per-profile proxies, environment checks, and Claude / Codex / Gemini terminal launchers.

    nodecloakRust★30

  • B awsegress Draws where data flows out of AWS onto a single diagram, saving effort for cost or compliance reviews; the idea is simple and effective, but the repo is under two days old and it's unclear which services are covered. #DevOps#Observability#Privacy#aws

    AWS Egress map for you to understand where your data is going.

    petricbrankoPython★12

  • B video-prediction-policy-2 The second-generation official implementation of a robot video-prediction policy, a hot direction, but the description is only a title with no word on whether weights are released or how to run it. VLA researchers may want to bookmark it, but ordinary developers have nothing to run after installing. #ML Research#Robotics#AI

    Official implementation of Video Prediction Policy 2: Predict Better, Act Better

    roboteraxPython★19

  • B sweety-image-art-direction An explicitly triggered image art-direction skill that emphasizes preserving style, clarifying reference-image usage, and checking output each round. The workflow is clearer than most prompt collections, but it is still prompt orchestration, replaceable by the reference-image features in drawing tools. #Agent Skills#Image Generation#Prompting#AI

    Image art direction: preserve the art style and subject, clarify the purpose of reference images, organize prompts and check the generated images round by round. Standalone, explicit-only AI art direc…

    threerocksLanguage unknown★11

  • B ZRList A virtual list/grid component for Unity uGUI supporting dynamic sizes, multiple prefabs, collapsible trees, and nested scrolling, real pain points for Unity developers. Paid options like EnhancedScroller exist, so being free is its edge, but it is just another implementation lacking documentation. #Game Dev#UI Design#unity

    Unity uGUI virtualized list and grid components with dynamic sizing, multiple item prefabs, collapsible trees, nested scrolling, and animated navigation.

    zhanglinshuia-codeC#★13

  • B macos-on-ios-public Using retag plus a UIKit compatibility layer to bring unmodified Apple Silicon macOS games to iPad, an approach that is fresh among recent similar attempts. For now it is more a tech demo that cannot be installed or used well, suited to those who want to see how the bypass works. #iOS#macOS#Reverse Engineering#Games

    Run unmodified native Apple-silicon macOS games on iPad (retag + UIKit shims)

    zorrobyteObjective-C★9

  • B tardigrade It adds persistence to agents, recovering after the process is killed via Durable Objects and Artifacts. A recoverable agent runtime is an interesting direction, but the description is too slogan-like and doesn't clarify recovery semantics or integration cost. #AI Agents#LLMs#cloudflare-workers

    An agent you can kill, and it comes back. Pi Durable + Cloudflare Durable Objects + Artifacts.

    acoyfellowTypeScript★12

  • B rebalance-review A browser-based ETF allocation comparison terminal doing risk attribution and historical replay, aimed at paid tools like Portfolio Visualizer. But wording such as "permission-aware reports" looks template-generated, and the source of 44 stars in 1.9 days is questionable. #Fintech#Data Analysis#Web Frontend

    Compare ETF allocations with a browser-based research terminal: risk attribution, historical replay, cost scenarios and permission-aware reports.

    yz3639-gifTypeScript★44

COrdinary50
  • C founder-skill Eleven MIT skills that validate startup ideas via a Hormozi/Thiel/Jobs board, a CFO and hundred-buyer agents, riding early agent-skill momentum to 144 stars in 2 days. Monetization could be a skill catalog subscription or consulting funnel; the risk is uneven skill quality and being absorbed by built-in general-agent capabilities. #Agent Skills#Prompting#Marketing#LLMs

    Eleven free Claude skills that test a business before you launch it: a board trained on Hormozi, Thiel and Jobs, a marketing director, a CFO, and a consumer panel of 100 buyer agents. Free, MIT.

    JakeschincariolPython★144

  • C multicoin-address-validator Validates address formats for Bitcoin, Litecoin, Ethereum and more — a must-have pre-check for wallet and exchange deposits, and 231 stars in 0.3 days shows real demand. But it's a republish of an existing package with low novelty, and the validation logic is easily absorbed by native chain libraries, so it's a long-tail standard dependency rather than a business. #Crypto#Fintech#SDKs

    Useful library for validation of Bitcoin, Litecoin, Ethereum and other cryptocoin addresses

    Nonco-OrganizationJavaScript★231

  • C sentinel-backend FastAPI handles Stellar account screening, Soroban event monitoring and network health, the actual service surface of this compliance product. Three repos each hit 130 stars the same day, showing promotion but no customer signal. On-chain AML has real paying customers, and it could move to subscriptions if it integrates exchange and custodian APIs. #Blockchain#Monitoring#APIs

    FastAPI service for Stellar account screening, Soroban event monitoring, and network health.

    Stellar-SentinelPython★130

  • C ai-newtab Claude generates a new tab page daily from browsing history, and 149 stars in 0.3 days shows good momentum and appeal. But it requires reading all browsing history, a privacy boundary hard to explain to users, and new-tab ad monetization conflicts with the no-tracking narrative, making the business logic self-contradictory and hard to become a standard dependency. #Browser Extensions#Privacy#LLMs

    Claude generates a daily new tab page based on your browser history.

    ThariqSTypeScript★149

  • C sentinel-contracts The Soroban contracts handle authorization flags, thresholds and latest-flag records, the trust layer of on-chain screening. Three repos from the same organization each got about 130 stars the same day, so momentum exists but looks like a coordinated release lacking independent-user evidence. The contract layer is too thin to stand alone and needs backend customer data to back it up. #Blockchain#SDKs#Fintech

    Soroban contract for authorized Stellar Sentinel flags, configurable thresholds, and latest-flag records.

    Stellar-SentinelRust★131

  • C sentinel-frontend A Stellar screening dashboard showing flag events and network health, a companion UI rather than a standalone product. It got the same star level on the same day as two sibling repos, so momentum comes from a coordinated release, not organic adoption. The frontend has no standard dependency attributes, and its commercial value depends entirely on the backend landing customers first. #Blockchain#Web Frontend#Monitoring

    Stellar Sentinel dashboard for Stellar account screening, network health, and Soroban contract flag events.

    Stellar-SentinelTypeScript★130

  • C ema-lightning-api A self-hosted Turkish TTS API that runs as a single container and supports streaming synthesis. It saves teams needing tr-TR voices some work, but similar wrappers are plentiful and the project doesn't state its model source or quality data. #Speech#Self-Hosted#APIs

    Self-hosted Turkish text-to-speech API. One container, full and streaming synthesis.

    bariskisirPython★9

  • C ste-zht A skill that makes agents report in Traditional Chinese while simplifying technical English per ASD-STE100 principles. The scenario is very narrow but clearly defined; technical-writing teams can reference it, with limited value beyond that. #Agent Skills#Prompting#Documents

    Agent skill: report results to the user in Traditional Chinese following ASD-STE100 principles

    dualfaceLanguage unknown★12

  • C FileBrowserMobile An Android WebView client for self-hosted FileBrowser Quantum. The need is real but the implementation is just a shell; those using FileBrowser who want quick phone access can try it, though technically there's no increment. #Android#Self-Hosted#filebrowser

    Android WebView client for self-hosted FileBrowser Quantum

    durdoomKotlin★12

  • C jumper-replica Reverse-engineers a product's geometry data and provides STL, STEP, Blender exploded views and a test-fit 3MF, while stating the physical object is unverified; the honesty is a plus, but it only serves the few who want to make their own replacement parts. #3D#Reverse Engineering#Hardware

    Independent Jumper geometry recovery and assembly references: STL, faceted STEP, Blender exploded views, and pilot-fit 3MF. Physical replica not yet validated.

    manifoldorPython★38

  • C ste-prompt Turns the aviation industry's Simplified Technical English STE100 rules into a coding agent plugin, using controlled language to cut ambiguity; the idea is concrete and rare, but the deliverable is prompts, making effectiveness hard to verify. #Prompting#Agent Skills#AI Agents#claude

    ASD-STE100 Simplified Technical English as a plugin for Claude Code, Codex & coding agents — clearer, unambiguous LLM output

    menonpgLanguage unknown★13

  • C 1401 Focuses on 'running macOS on your existing PC', addressing developers' pain of wanting a low-cost macOS environment, with 137 stars showing notable momentum. But it claims it will never be sold, closing off any commercial path; more importantly, it violates Apple's EULA and could face legal action anytime, making it hard to become a standard CI/CD dependency. #macOS#Operating Systems#Hardware

    1401: macOS on the PC you already own. Free, never for sale.

    nullmothPython★137

  • C pl-flow A two-stage TTS with reference audio, outputting 32kHz mono, with the method explained more clearly than most peers; but open-source TTS is already crowded, and the description alone shows no perceptible advantage. #Speech#ML Research#AI

    PL-Flow: a two-stage, reference-conditioned text-to-speech model producing 32 kHz mono audio.

    rcell233Python★16

  • C playgta5 Open-sources playgta5.com, gaining 703 stars in 1.6 days, extremely strong momentum, but the traffic relies almost entirely on the copyrighted GTA5 IP. Technically it's just a browser-side shell/streaming, and commercially it can't be legally monetized or become a dependency, facing takedown by Take-Two at any time. It's a traffic-driven repo, not an investable project. #Games#Web Frontend#Downloaders

    A source code for playgta5.com.

    shadany7824JavaScript★703

  • C factory-seat A career mode for Le Mans Ultimate, climbing from LMGT3 up to Hypercar, honestly positioned as an unofficial fan project. Useful only to players of that racing sim, and at just over a day old its completeness is unclear. #Games#Game Dev#le-mans-ultimate

    Factory Seat: a career mode for Le Mans Ultimate. Multi-season driver careers from LMGT3 to Hypercar. Unofficial fan project.

    factoryseatlmuappC#★10

  • C formulaire Reworks Gridforms' idea of handling form layout with a single stylesheet, offered as plain HTML or Base UI-based React components; the judgment is sound, but it solves a small problem and similar options are plentiful. #Web Frontend#UI Design#react

    A modern take on FormKeep's Gridforms: one stylesheet for plain HTML, or React components built on Base UI.

    mattrothenbergTypeScript★10

  • C riskbound Turns volatility targeting and a 200-day moving average into explicit position rules, with a forward record and an 'honest test' framework, phrased with restraint; but it's just an implementation of a few rules, offering limited insight for quant readers. #Fintech#Data Analysis#Testing

    Position-sizing rules with explicit evidence boundaries (volatility targeting + 200-day SMA), forward log, and an honest-testing harness.

    naplesbluePython★9

  • C ema-lightning-web Runs a Turkish TTS model in the browser via ONNX, offering both WebGPU and WASM paths; clever engineering, but the audience is very narrow, and single-digit stars show nobody is really using it yet. #Speech#WebAssembly#GPU#onnx

    Türkçe TTS modeli EMA Lightning'in ONNX ile WebGPU ve WASM üzerinde tamamen tarayıcıda çalışan sürümü.

    ozcancelikJavaScript★9

  • C esp32-cpu-ram-gauges Drives physical needle gauges with an ESP32-C3 to show PC CPU/memory usage, a small hardware toy described honestly without hype. Similar desktop monitoring ornaments already exist, so unless you happen to want to solder one, the appeal is limited. #Hardware#Embedded#IoT

    Physical CPU/RAM Gauges powered by ESP32-C3

    squawkiC++★9

  • C resimind A neuro-symbolic agent with verifiable memory and optional Lean integration for validation, an interesting direction; but the repo is less than a day old with no runnable examples or install instructions, so actual capability is unclear and it currently reads like a research draft. #AI Agents#ML Research#LLMs

    Neuro-symbolic agents with verified memory. Use what you know; call the model when needed. Self-correction, optional Lean checks and reusable rules.

    1105216375-altPython★9

  • C promiscope-frontend It builds an on-chain accountability dashboard from community commitments, milestones, progress and evidence links, with a public-good social angle but a narrow real-world use case. Its 130 stars in one day match other projects in the same ecosystem, suggesting batch releases; with no paying customers or growth evidence, its commercialization path is the weakest in this batch. #Blockchain#Collaboration#Crypto

    Community project accountability on Stellar: commitments, milestones, progress updates, evidence links, and community responses.

    Stellar-PromiscopeTypeScript★130

  • C bitcoinix It packages a Bitcoin full node and NixOS modules with Nix, emphasizing secure configuration. That's convenient for Nix users, but nix-bitcoin has done the same for years, and this states neither differences nor validation results. #Crypto#Self-Hosted#nix

    A collection of Nix packages and NixOS modules for easily installing full-featured Bitcoin nodes with an emphasis on security.

    bitcoinixNix★9

  • C wp2shell Chains two CVEs into a WordPress RCE, gaining 139 stars in 1.3 days; it has patch-validation value for defenders, but a public exploit can't be commercialized or become a dependency, disclosure channels like WPScan already exist, and it carries abuse and legal risk, so it's only worth short-term tracking. #Exploits#App Security#Network Security

    CVE-2026-63030 + CVE-2026-60137 exploit RCE chain

    manpisetsuPython★139

  • C zcode-claim-9router Automatically claims z.ai's free Start plan and hooks GLM into 9Router by reverse-engineering the client protocol to grab quota. Scripts like this break with any upstream change, abuse the free policy, and are technically one-off adaptations, not worth attention. #LLMs#Reverse Engineering#APIs#zcode

    Automate the ZCode free Start-Plan claim (z.ai token/cookies -> ZCode JWT -> Start Plan) and wire GLM into 9Router as a glm provider pool. Web UI console + CLI. Reverse-engineered from ZCode 3.14.4.

    0xgetzPython★29

  • C MosaicDiff Mosaic restoration using MiniMax H3 with a custom LoRA, a narrow image-restoration application whose real uses mostly point to adult content. Beyond model calls there are no method details, and it isn't suitable to open at work. #Image Generation#Adult Content#AI

    Restore mosaic using MiniMax H3 and custom lora

    dotaku22Python★26

  • C heartbeat It adds scheduled heartbeats to Claude Code so the agent proactively reaches people via Telegram. The idea is clear but the implementation thin; cron plus a webhook would replace most of it. #AI Agents#Automation#Productivity#claude

    Let your AI reach out first — fixed or self-scheduled heartbeats for Telegram + Claude Code

    leiiiyurongPython★10

  • C AgentGarten 'A code world for evolving agents' picked up dozens of stars in a day, but the description stays conceptual, with no environment interface, training pipeline or way to run it — for now just a research manifesto. #AI Agents#ML Research#LLMs

    AgentGarten: Code Worlds for Evolving Agents

    MirroS-LabPython★50

  • C qqpet A nostalgic single-player QQ Pet recreation with AI, an extremely narrow audience; the one-sentence description doesn't say how faithful the recreation is or how to run it, so value is hard to judge outside nostalgic players. #Games#Game Porting#AI#tencent

    QQ Pet single-player remake (AI)

    AnkioTomasTypeScript★12

  • C yippee-kids-chores-card A children's chore star-reward card for Home Assistant, a specific and very small scenario, essentially a hobbyist's self-use component with almost no transfer value beyond families also using HA with kids. #UI Design#Productivity#home-assistant

    Kinder-Aufgaben mit Sternen und Belohnungen für Home Assistant

    Chrism1412Language unknown★13

  • C x2Stock An A-share research workbench envisioning an independent data foundation, multi-agent research and quantitative validation, but the author states it's 'currently at the overall design document stage' — nothing runs yet, so visiting now just means reading architecture diagrams. #Fintech#AI Agents#Data Analysis

    An open-source A-share research workbench: independent data foundation, multi-Agent research and quantitative validation. Currently at the overall design document stage.

    MrLaoGeJavaScript★13

  • C ha-security A Home Assistant integration adding security visibility, a reasonable idea, but the description stops at 'improve visibility' without saying what data it collects, how to deploy it, or how it works with existing alerts, so feasibility can't be judged yet. #Privacy#Self-Hosted#home-assistant

    An integration to add visibility and understanding to the security around Home Assistant

    Anon666333Python★11

  • C opencode-free-9router Wraps OpenCode Zen's free quota as an OpenAI-compatible provider for 9Router, essentially a shortcut that bypasses registration and quota limits; it breaks as soon as the upstream changes and has no long-term value, not worth your time. #LLMs#APIs#opencode#9router

    Expose the OpenCode Zen free-tier models as one OpenAI-compatible provider for 9Router. No OpenCode account required. Zero dependencies.

    0xgetzJavaScript★25

  • C Sideport It adds a download APK button to Google Play pages, with six third-party sources and Morphe patch lookup. Patch lookup pushes it toward ad-removal and unlocking grey areas, and source reliability isn't its responsibility. #Android#Downloaders#Crack#Browser Extensions

    Userscript: a Get APK button on every Google Play app page — direct downloads from 6 sources, live availability checks, Morphe patch finder. No tracking.

    heval99JavaScript★15

  • C verifiedhandles-mcp It connects an identity-verification service into MCP, with client configs and a registry entry. Engineering-wise it's tidy, but its value depends entirely on that service—an integration most people won't need. #MCP#AI Agents#APIs

    Connect an AI agent to Verified Handles: setup for each client, the MCP Registry listing, a Claude Desktop extension.

    jameshaworthcsJavaScript★30

  • C Autotrader An automated trading script using Kronos for market forecasting, portfolio construction and paper trading, running on GitHub Actions. It sounds complete but has no backtest evidence or risk disclosure, and trading projects usually stop at the demo stage — worth a look to learn agent orchestration, not as a usable strategy. #Fintech#AI Agents#Automation

    Autonomous AI trading system powered by Kronos. Forecasts markets, builds portfolios and paper-trades automatically with GitHub Actions.

    PkLavcPython★20

  • C claude-mods It calls itself a mod collection for Claude Code but never lists which mods, how to install them, or what behavior they change. Collection projects suffer most from such one-line descriptions, and sixty-plus stars in a day says nothing about quality. #LLMs#Plugins#CLI

    A collection of claude code mods

    TickloopTypeScript★62

  • C zapret-discord Essentially a Discord-specific configuration package for zapret, with only a Russian slogan and no parameters, ISP applicability or differences from upstream. It may help immediately for those DPI-throttled in Russia, but it's unclear whether it's just a hotspot-riding repackage. #Networking#Privacy#discord

    An updated version for Discord under current conditions. More reliable, faster, more stable

    echoillusionistcurlBatchfile★50

  • C power-agent A minimal agent harness written in pure PowerShell, like a technical fun experiment; beyond running on Windows, it doesn't say how it beats existing harnesses, and few will actually install and use it. #AI Agents#CLI#Windows#powershell

    A minimal harness built entirely in PowerShell

    nakasyouPowerShell★9

  • C cardlane A bilingual gift-card marketplace on the Base Sepolia testnet with encrypted delivery, on-chain escrow and dispute windows. Plenty of concepts stacked up, but it's still a testnet demo and touches neither real settlement nor compliance. #Crypto#Blockchain#E-commerce#Fintech

    Bilingual gift card marketplace with encrypted delivery, onchain escrow, 24-hour appeal replies and 48-hour settlement on Base Sepolia.

    CryptoxiaoxiangTypeScript★16

  • C mama-skill An agent skill sold on the joke of 'letting AI be your mom' — fun over useful, with no description of concrete abilities, triggers or use cases, so it's just a meme. #Agent Skills#Prompting#AI Agents

    The only way to prevent humans from being wiped out by AI is to make AI a mother. — Geoffrey Hinton

    KoriIkuLanguage unknown★28

  • C margarethamilton.dev A scroll-narrative tribute page about Margaret Hamilton, who led Apollo lunar software; the production quality is worth a look, but it's a personal work and content page, not a usable tool. #Web Frontend#Animation#UI Design

    A scroll-driven tribute to Margaret Hamilton (1936–2026), who led the Apollo 11 flight software and gave software engineering its name.

    midudevAstro★25

  • C dsh-ops It offers bash, PowerShell 7 and Rust tools to save tokens for dsh on Windows, but never explains what dsh is, how to integrate it, or how much it saves. Outside readers can't judge its value, so it counts as half-finished. #Windows#CLI#LLMs

    Bash, PowerShell 7, and Rust-based tools for dsh on Windows to cut token usage. / High-performance bash, PowerShell 7 and Rust tools for dsh on Windows to reduce token consumption

    T-AutoJavaScript★10

  • C gluon One line: a coding agent control platform, without saying how it connects or what it controls. Several such orchestration layers have appeared this week, with no distinguishing technical point. #AI Agents#LLMs#CLI

    Control platform for coding agents

    fidiproTypeScript★38

  • C EndCraft It claims a native rendering bridge between Minecraft and Arknights: Endfield, but is only a prototype. Pairing two IPs looks more like an experimental gimmick, and the description doesn't show what running it would look like. #Games#3D#Game Dev#minecraft

    Experimental Minecraft × Arknights Endfield bridge: native scene rendering, Minecraft character and building prototype.

    guchang233C++★9

  • C free-ai-api A list aggregating free API quotas and trial allowances from various providers, a content-type resource that goes stale within weeks and has nothing runnable; useful only as a temporary lookup table, not worth a dedicated visit. #LLMs#APIs#Knowledge Base

    Free AI API tiers and trial quotas in 2026: Gemini, Groq, Cerebras, OpenRouter, Zhipu and more; maintained comparison and sources

    A2agent-aiLanguage unknown★12

  • C effect-cloudflare-foldkit-template It packages Effect, Foldkit, Bun and Alchemy into an organization template, convenient for teams on that stack. But templates inherently lack a problem statement, so readers can only trust the author's choices. #Build#CI#Web Frontend

    Reusable organization, tooling, and guardrails for Effect 4, Foldkit, Bun/mise, and Alchemy 2 on Cloudflare

    just-be-devTypeScript★38

  • C whenever From a big-company incubator, "AI workflow automation for the agentic era" states only that vision, with no explanation of the orchestration model, triggers, or how it differs from existing workflow engines. The name is grand, but there is too little to read. #AI#AI Agents#Automation

    AI Workflow Automation for the agentic era

    wix-incubatorTypeScript★18

  • C AppMarket States outright that it's a fork of Uevo001/AppMarket, with an F-Droid repo and signature-verified updates. Being a fork means it isn't a new project, and its value depends on what changed versus upstream, which the description doesn't explain. #Android#Self-Hosted#f-droid

    Fork of Uevo001/AppMarket - KMP cross-platform app store (Android/Desktop) with F-Droid source and signature-gated updates

    BeiChen-CNKotlin★17

  • C ps5-fpkg-converter A converter that turns fake PS5 packages into a bootable format for jailbroken consoles, meaningful only for already-jailbroken machines and part of the cracking chain. There's engineering effort behind it, but the purpose itself isn't suitable to recommend to ordinary readers. #Crack#Piracy#Reverse Engineering

    Convert PS5 fake packages (.pkg / FPKG) into launchable formats (folder dump, .ffpfs, .ffpfsc) for jailbroken consoles. Windows CLI + GUI, built on ps5-ultrapack.

    sub0x00001Python★12

  • C open-intelligent-ui From a team with background, but the empty repo description leaves it unclear whether this is a component library, protocol, or rendering approach. 129 stars in 0.8 days shows high interest, yet readers find nothing inside; wait for documentation. #AI#Web Frontend#UI Design

    thesysdevTypeScript★129

DSkip73
  • D promiscope-contracts The repo says it is kept only for legacy compatibility, with the core design on another branch, indicating the contract layer is transitional and current code has limited value. Its 129 stars appeared the same day as the frontend, driven by the overall release rather than itself. On-chain accountability demand is unproven, so skip it and wait for the new contracts. #Blockchain#Crypto#APIs

    Promiscope Soroban contracts retained for legacy workflows; privacy-safe accountability contract design is tracked separately.

    Stellar-PromiscopeRust★129

  • D RecursiveGameCreator A buzzword-stacked description like 'agent product-experience-oriented game harness' gives no sense of the actual output or how to start — a repo where vision outruns implementation. #AI Agents#Game Dev#Automation

    Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness

    IMBALDYPython★31

  • D StdXRobot A Telegram group-management and song-request bot whose description piles fancy adjectives, with features heavily overlapping existing off-the-shelf bots. It says nothing about deployment or how it differs, making it a feature-stuffed toy. #Social#Automation#Self-Hosted

    ᴛʜᴇ ᴍᴏsᴛ ᴩᴏᴡᴇʀғᴜʟ ᴛᴇʟᴇɢʀᴀᴍ ɢʀᴏᴜᴩ ᴍᴀɴᴀɢᴇᴍᴇɴᴛ ➕ ᴍᴜsɪᴄ ᴍᴀɴᴀɢᴇᴍᴇɴᴛ ʙᴏᴛ ᴀɴᴅ ɪ ʜᴀᴠᴇ sᴏᴍᴇ ᴀᴡᴇsᴏᴍᴇ ᴀɴᴅ ᴜsᴇғᴜʟ ғᴇᴀᴛᴜʀᴇs.

    STD-DEEPANSHULanguage unknown★28

  • D s6e8-lookup-transformer A personal solution to a Kaggle competition, honestly stating its ranking and citations. But anyone not in that competition gets no reusable value; it's essentially a record of a contest assignment. #ML Research#Data Analysis#Benchmarks

    My Lookup-Transformer solution for Kaggle S6E8: public LB ROC AUC 0.97041. Adopted in a 14th-place solution; 11 verified Kaggle references.

    SpydonnyJupyter Notebook★10

  • D dailileida It tracks star and store-rating growth for over sixty proxy clients and ranks them, using public metrics and staying lightweight. Somewhat useful for choosing a client, but it's essentially an aggregation page without its own judgment. #Networking#Data Analysis#Privacy

    Daily tracking of 60+ proxy clients including Clash Verge Rev, v2rayN, FlClash, Hiddify and Shadowrocket, ranked by GitHub Star and store rating growth, highlighting fast-growing projects.

    guiforcoresHTML★15

  • D hongguo-drama-mac A Mac short-drama player and downloader scraping Hongguo and Xifan, with specifics like episode merging. But the content's copyright is clear-cut; it's a piracy scraping tool and not recommended. #Piracy#Video#Downloaders#macOS

    Open-source Mac short-drama player and downloader · Hongguo / Xifan dual sources · episode download, continuous playback and per-episode merging · Apple Silicon / macOS 13+ · GPL-3.0

    jackotomJavaScript★12

  • D LMS A multi-tenant self-hosted LMS listing course, instructor and student sides; it directly overlaps with Moodle and Open edX, and gaining just over ten stars a day hardly proves these modules actually run. #Self-Hosted#Learning#Collaboration

    Self-hosted, multi-tenant learning management system (LMS) for training institutes and teams: course builder, trainer and student portals, progress tracking. Django + React. Apache-2.0.

    pacewisdomsolutionsTypeScript★10

  • D OpenSpawnAI Orchestrates multiple models into a 'company that runs itself', but it's all vision with no operational detail: no install steps, no model hookup, no idea what you'd see when running. Agent-orchestration frameworks are already plentiful, and this one offers neither differentiation nor use cases — classic big-word packaging. #AI Agents#LLMs#Automation

    Build a company that runs itself. AI Models are your workers - connect them in departments and create self-running companies. Windows, Mac, and Linux.

    OpenSpawnAITypeScript★9

  • D PlayerVoice-GameAnalysis-CN The description is all empty phrases like traceable facts, community evidence and structured insight, with no sense of inputs or outputs. Game sentiment analysis already has many solutions; this offers wording, not method. #AI#LLMs#Data Analysis#Games

    Evidence-backed multilingual game research pipeline for players, creators and analysts, with traceable facts, community evidence and structured insights.

    feng141368-cyberPython★17

  • D battery-saver-windows-upd A small Windows tool for low-battery power saving and battery health, claiming no OEM driver dependency. But the repo has no language or implementation notes, and both author and repo names contain suspicious 'crack' wording, so a system-level tool of unknown origin is not recommended for installation. #Windows#Hardware#Performance

    Power Saver at low battery, battery health tracking. Works on Dell, HP, Lenovo, Asus - no OEM drivers.

    sheenbreakercrackLanguage unknown★13

  • D ffxi-2016-ps2 A fan project for the 2016 PS2 version of Final Fantasy XI aimed at a tiny nostalgia niche; the description never says whether it ships a client, server or patch, so outsiders can't judge usability. #Games#Game Porting#ffxi

    FFXI 2016 PS2 Edition. Fan Project by Habex.

    HAB3XLua★11

  • D oss-scanner The name sits under anthropics, but there is no description, so you can't see what it scans, how to use it, or what it depends on. Whoever is behind it, an undocumented repo shouldn't take up ten minutes of readers' time. #App Security#Code Quality#Secrets

    anthropicsPython★43

  • D compositor-cn A Simplified Chinese localization of MIT's macOS image editor Compositor, with no functional changes; it saves Chinese users a step, but such localization repos struggle to track upstream and offer nothing new. #Images#macOS#Editors

    Compositor Simplified Chinese localization — free open-source macOS image editor (upstream: robbietilton/Compositor, MIT)

    mainanamisaSwift★9

  • D FoxyVPN An unofficial Firefox VPN Windows client billed as 'subscription-free' — in effect a way around paid subscriptions, with unverifiable origin and security. Not recommended for installation. #Networking#Privacy#Windows

    Subscription-free unofficial Firefox VPN client for Windows.

    M-RTZ1Dart★13

  • D Cybersecurity-Mastery A collection of security learning materials with no code and no tools, essentially a resource-list content repo. Beginners wanting a structured path can use it as a table of contents, but it isn't a new project worth ten minutes today. #Learning#App Security#Network Security

    STD-DEEPANSHULanguage unknown★29

  • D call_call An experience-sharing write-up on call-quality details with recommended reference libraries — essentially a notes collection with no installable product, so readers only get scattered tips. #Speech#Web Frontend#Documents

    Sharing some detail work done on a personal frontend to improve the call experience, plus recommended reference libraries

    KKarsylineSwift★14

  • D ai-business-skills Eight skill files teaching agents to build and sell products—a prompt and skill collection with no code or verifiable method. Similar skills repos have appeared in droves this week, leaving readers unable to choose. #Agent Skills#AI Agents#Marketing#Prompting

    Eight skill files for building and selling a product with an AI agent

    echris6Language unknown★12

  • D ps5-atmosphere-installer It installs PS5 games from local SMB/FTP, actually used with jailbroken firmware, and the Atmosphere in its name is the Switch one. The tool is thin and clearly in piracy's grey zone. #Piracy#Games#Downloaders#ps5

    install ps5 games from your local smb / ftp storage

    flex36tyC★9

  • D cs2-egg-checker-server The description is just one line about joining and checking pet info in the console, with no clue what it does, who it serves or how to use it; there isn't enough information to judge whether it's worth opening. #Games#APIs#cs2

    Join and check your console for your pet information

    BeepIslaC++★10

  • D skills Another engineering and design agent skills collection—a few markdown files with no scenario trade-offs or usage results. Similar repos are appearing at several per day. #Agent Skills#AI Agents#Prompting

    Agent skills for engineering and design

    ishuagrawalJavaScript★19

  • D ARTEX Self-labeled as a source backup of Autumn-27/ARTEX, keeping only the original commit history and license — a mirror repo that produces nothing new. #Backup#AI#Productivity

    ARTEX source backup: based on Autumn-27/ARTEX v0.3.15, preserving the original commit history and AGPL-3.0 license.

    HinlnGo★53

  • D Qwen3.8-Flash-Dual-DGX-Sparks-TensorFold Claims to run Qwen3.8-Flash-Next plus TensorFold on two DGX Sparks, but neither the model name nor the scheme maps to public information — more like a placeholder repo chasing hype. #LLMs#GPU#ML Research#Hardware

    Qwen3.8-Flash-Next on two DGX Sparks with TensorFold

    MiaAI-LabShell★10

  • D Xeno-Executor A Roblox script executor touting keyless operation and speed, essentially a cheat tool. Beyond violating platform terms, it offers no technical content worth learning from and often carries security risks. #Game Cheats#Games#roblox

    Xeno Executor - best Free Roblox script executor in 2026. Lightning-fast execution, external script support, and a simple GUI - all completely keyless.

    Xeno-oniHTML★46

  • D 3d-card-game-showcase Core scripts and UI Toolkit demos for a Unity 3D card game, reading like a portfolio; there's no playable build or complete documentation, so learners only get fragments. #Games#Game Dev#3D

    Unity / C# 3D card game: core scripts, turn-based combat, map generation and UI Toolkit

    nocTia-O9C#★13

  • D gbos-vm Claims to virtually run 'Googlebook OS' on Apple Silicon, but no such system exists, and the description corresponds to no real product. With 52 stars in 1.3 days, it looks like name-baiting, so skip it. #macOS#virtualization

    Run Googlebook OS in a GPU-accelerated VM on Apple Silicon Macs

    skylartaylorPython★52

  • D zp12 The description reads "attempt to Sp12 on Fm-1 m-vave," an experiment in flashing firmware to the M-Vave FM-1 synthesizer. The target is extremely niche and even the changes are unclear, so only owners of the same hardware can judge its value. #Embedded#Hardware#Reverse Engineering

    attempt to Sp12 on Fm-1 m-vave

    zvensonC★12

  • D Haimiya-Mio The second version of a personal portfolio site — a personal showcase repo with no reusable library or tool, so of limited value to others. #Web Frontend#UI Design

    My Portfolio v2 version

    Itz-AnyaTypeScript★77

  • D input-optimization-toolmenu The title is stuffed with keywords like INPUT/LAG/BOOST and claims lower latency and better game performance, yet there is no code explanation or verifiable method. Such "optimization tools" are usually registry tweaks plus placebo, often bundled with risk, and not worth the time. #Games#Performance#Windows

    INPUT | LAG | FIX | MOUSE Boost Gaming Performance | Lower Latency |

    QuantumSupplierLanguage unknown★16

  • D Python-Mastery A Python study-notes repo by the same author, with an empty description and pure content. It may help beginners a little, but it offers no new approach and nothing runnable, so it isn't an open-source project worth following. #Learning

    STD-DEEPANSHULanguage unknown★29

  • D lsc A League of Legends skin changer that swaps skins by modifying client resources, with a title full of emoji and year keywords. Such tools violate game terms and often carry security risks, and there is no technical content worth learning from. #Game Cheats#Games#Crack

    🔥 Lol-Skin-Changer-2026

    SteersmanReleaseHTML★50

  • D simhubforum A 'SimHub full bundle download' that packages and redistributes third-party software; with no version notes or licensing information, it's a resource-site-style repository. #Piracy#Downloaders#Games

    simhub full bundle download

    maow318Language unknown★36

  • D Kinlock-Org.github.io The docs-site source for some project, explaining only how four repos fit together — documentation rather than a tool, with no standalone value apart from the upstream code. #Documents#Knowledge Base#Privacy

    Kinlock documentation site (GitHub Pages): how the 4 repos fit together

    Kinlock-OrgHTML★11

  • D Transformer_study Transformer study notes, a personal reading record with no tooling, no new conclusions and no specific problem solved. #Learning#LLMs#ML Research

    chenjianyong909Jupyter Notebook★11

  • D TurboFPS A Windows speed-up repo whose only description is 'open-source tool, updated 2026-10-09' — vague and with a suspicious date, it fits the classic 'optimizer' pattern with no verifiable implementation details. #Windows#Performance#Games

    TurboFPS - open source utility, updated 2026-10-09

    CobraRegenerateC++★17

  • D wardogs-optimizer An emoji-laden Windows 'performance optimizer' claiming higher frame rates, lower latency and one-click tuning — a typical unverifiable system tweaker with exaggerated claims. Best avoided. #Windows#Performance#Games

    ⚡ Wardogs Optimizer — Advanced Windows performance tool. FPS boost, latency reduction, system tweaks, background process control. One-click optimization for competitive gaming.

    DirectorReleasePython★10

  • D video-production-skills No description, so one can only guess from the name that it's a collection of video-production skills, with no visible content or usage. Repos you can't tell what they do shouldn't take up readers' time. #Video#Agent Skills#AI

    Ucmp0001Python★24

  • D build-flutter-demo-app-using-react No description and no language tag, the name suggests a throwaway demo comparing frameworks. Nothing reusable is visible, and there are no conclusions to learn from. #Web Frontend#Learning#Android

    abdelrhman-mahrousLanguage unknown★9

  • D robotworld A Rust repo whose README description field is empty, with neither name nor content revealing what it does. Its purpose can't be judged and it can't be tried, and 30 stars in 1.8 days offers no verifiable support, so skip it for now. #Robotics#AI

    robotworldaiRust★30

  • D Pokemmo2Pro-Client The description is just "28887 client," with no way to tell whether it is a game client or server companion, and no gameplay or deployment notes. Similar private-server clients usually have very short lifespans and are not worth the time. #Games#Game Dev

    28887 client

    u123510Java★11

  • D Pokemmo2Pro-Server A server based on openmmo, with a description that is likewise only a one-line source note and no version, dependencies, or run instructions. It appears paired with the same author's client, but there is too little information to judge usability. #Games#Game Dev#Self-Hosted

    Based on openmmo

    u123510Java★15

  • D MEMO The repository has no description at all, so its features, purpose and how to run it can't be determined; per the rules, it isn't recommended. #AI#Productivity#Learning

    Cecilia384Python★20

  • D space_game A small ESP32 mini-game practice project, self-described as for the author's own amusement, lacking gameplay docs and reusable components — limited reference value for embedded developers. #Embedded#Games#Hardware

    SPACE GAME - Joguinho usando ESP32

    GabrielDev0904C++★9

  • D Cocos-ai-copilot No description and no language tag, so one can only guess from the name that it's an AI assistant for the Cocos engine. No usage and no scope are visible; repos you can't tell what they do are never recommended. #Game Dev#AI#Plugins

    Renshuxian-aiLanguage unknown★18

  • D viralwave-studio-guide A promotional guide for a paid social-media automation tool, covering features, plans and FAQ. It's essentially a marketing landing page rather than an open-source project, with no usable code. #Marketing#Content Writing#Social

    The complete guide to ViralWave Studio: put your social media content on autopilot. Setup walkthrough, features, brand voice, review workflow, AI agent integrations, plans, and FAQ.

    cporter202Language unknown★31

  • D nextjs-playground-seat-reservation A Next.js practice project, a seat-booking demo, with no description of what real problem it solves. Such playground repos appear daily, and unless you have a specific teaching need, there's no reason to open it. #Web Frontend#Learning

    rostyslav48JavaScript★11

  • D identity-v-newest-menu A collection of mod menus/tools for Identity V, a game-cheat project with a keyword-stuffed description and no technical content, ruled out immediately. #Game Cheats#Games#Reverse Engineering#identity-v

    Identity V modding,Identity V tools,Identity V customization,Identity V PC mods,Identity V utilities

    BridgedesinfernoLanguage unknown★10

  • D ibm-updated-ai The description is a templated 'AI startup blueprint generator' with no visible actual functionality or usability — it reads like an auto-generated project page. #AI#Marketing#Content Writing

    Powered by AI Startup Blueprint Generator built with Next.js, IBM watsonx.ai,

    MaintenanceCatalog96Language unknown★10

  • D fb-andgads-manager From the name alone it looks like a Facebook/Google ads management tool, but the candidate info doesn't explain its capabilities, usage or tech stack, so it's impossible to tell if it beats existing Ads API scripts; less than a day old with 10 stars is no evidence. It might help ad teams, but isn't worth ten minutes yet. #Marketing#Automation#APIs

    atombowderrickLanguage unknown★10

  • D fastapi-data-export There's no README, so one can only guess from the repo name that it exports FastAPI data; usage, supported formats and boundaries are all unstated, which amounts to no delivery documentation. #APIs#Databases#DevOps

    mazu-rokPython★10

  • D Player_Crack Cracks for players such as Xiaohuan Video and Hills, a tool for bypassing payment and licensing, which is outside our scope. There is nothing technical worth discussing either, so skip it. #Crack#Piracy#Video

    Xiaohuan Video, hills lite, hills for android

    xrsbLanguage unknown★16

  • D OpenBRender A repo with no description whose name hints at rendering, but nothing explains what it does or how to start, so it can't be judged. #3D#Images#AI

    GPTchatlyTypeScript★9

  • D TopazStarlightUI A repo with no description that looks from its name like the UI layer of some image software, but nothing says what it does or how to use it. #UI Design#Images#AI

    LoveRain1997Python★9

  • D electrum A new repo named electrum with no description and no stated relation to the Bitcoin wallet of the same name; the same org mass-creates repos, suggesting name squatting. Not worth considering. #Crypto#Fintech#Blockchain

    Nonco-OrganizationPython★66

  • D python-ecdsa Its description is word-for-word that of the well-known python-ecdsa library, and the repo drew dozens of stars one day after creation while the same org mass-creates repos — likely a same-name mirror or squat, not an original project. #Crypto#Blockchain#Secrets

    pure-python ECDSA signature/verification

    Nonco-OrganizationPython★72

  • D algofi-smart-contracts No description at all, yet it gained twenty-plus stars within hours of creation, and the content is a re-upload of existing DeFi contracts. It states no purpose and adds no value, a typical mirror/placeholder repo to skip. #Crypto#Blockchain#Fintech

    REFE-OrganizationPython★21

  • D dimension-adapters It appeared at the same time as two other repos from the same organization, with an empty description and pure re-upload, apparently a copy of a DeFi data adapter. Stars climbed to eighty-plus within hours, which doesn't look like organic growth, and there is nothing to see. #Crypto#Blockchain#Data Analysis

    REFE-OrganizationTypeScript★83

  • D peggedassets-server Another server-side repo with an empty description, apparently re-uploading an existing stablecoin data service. With no explanation and no changelog, it's impossible to tell what it does, and it doesn't constitute a new project. #Crypto#Blockchain#APIs

    REFE-OrganizationTypeScript★68

  • D red-team-cybersecurity-roadmap No README; the title is all there is—another red-team learning-path list. Such lists solve no specific problem, and readers already have more complete alternatives. #Learning#App Security#Network Security

    fawadqureshi007Language unknown★13

  • D recap-en-to-mm The repo has no description at all, and from the name one can only guess it's an English-to-Burmese summary converter, with no way to tell what it does or who it's for. A repo that can't be evaluated doesn't make today's list. #LLMs#Content Writing

    tharlaimarPython★9

  • D MediaFinder A Python repository with no description; the name hints at media search or download, but what it does and how to use it are unknown. Repositories whose purpose cannot be determined do not enter our recommendations. #Downloaders#Data Analysis

    xToshiroPython★12

  • D front-2sem-aula-17 A front-end course assignment repo with no description, showing no goal, output or reusable content; it's classroom practice with no reference value for readers of this list. #Learning#Web Frontend#Collaboration

    1TDSPJ-26TypeScript★17

  • D synthesia-newest It lifts Synthesia's marketing copy as its description, claiming 230+ digital avatars and 140 languages, with no implementation notes at all. A typical empty shell riding a paid product; opening it only wastes time. #Video#AI#Piracy

    studio-quality AI videos with 230+ avatars and 140 languages - including custom avatars from your own footage

    RogueAccountantBurstLanguage unknown★10

  • D kinetikern2 The description is a single word, with no explanation or usage, so it's unclear what it does. A repository that can't be judged isn't recommended. #AI#ML Research#Science

    Spacing

    bghryctPython★12

  • D leonardo-mod-menu Titled as a Leonardo AI advanced automation suite, it's actually an empty shell page riding that drawing product's name, with no usable implementation—a grey-market traffic-farming repo. #Crack#Image Generation#Scams

    Leonardo AI Premium Automation Suite 2026: Ultimate Creative AI Toolkit & Workflow Generator

    chartostrichcondenseLanguage unknown★10

  • D minecraft Titled 2026 crack client with flame emojis everywhere, the repo is an HTML shell—a typical Minecraft-keyword star farm with no technical content to assess. #Game Cheats#Games#minecraft

    🔥 Minecraft-Hacked-Client-2026

    flintserviceproxyHTML★50

  • D supermarket-management-system-showcase A Spring Boot plus Vue supermarket management system, with sanitized source and database schema, a typical coursework project; similar templates are everywhere and it solves nothing new. #Learning#Databases#Web Frontend

    Spring Boot + Vue supermarket management system: desensitized frontend and backend source code and database schema

    nocTia-O9Vue★10

  • D mastering-software-testing Student materials accompanying a course, with no explanation of which testing techniques are covered and no runnable examples; it offers almost no new information for readers of this list. #Learning#Testing#Code Quality

    Mastering Software Testing — Student Materials

    qursaanLanguage unknown★12

  • D Valorant-External-Undetected-2026 An external aimbot/ESP cheat for Valorant claiming to be undetected — a game-cheating project, ruled out outright. #Game Cheats#Games#valorant

    This Valorant cheat enhances gameplay with smooth aim, enemy visibility, instant shooting, and guaranteed hits, providing players with a tactical edge without detection for an improved experience

    MatterofpergolaC++★66

  • D rblm A Roblox mod menu — a game-cheating repo whose description is only tag-like keywords and emoji, with no technical content. #Game Cheats#Games#roblox

    🔥 Roblox-Mod-Menu-Free-2026

    ModernJellyfishHTML★44

  • D rbl A Roblox script executor, also a cheating project, with a description that is pure keyword stuffing and no implementation details to inspect. #Game Cheats#Games#roblox

    🔥 Roblox-Scripts-Executor-2026

    Mountainnulift12HTML★44

  • D monitorOS The README is empty, not even saying what it monitors or which platform it runs on, and 25 stars are unexplained. The name looks like a toy OS, not something you install and use. #Monitoring#Linux#Operating Systems

    eriskhanC★25

  • D Solana-Drainer-Tool The description explicitly states it automatically drains Solana wallets, supports Phantom/Backpack and uses covert routing, making it a phishing wallet-stealer; regardless of the technical implementation, it should not be recommended to readers. #Scams#Crypto#Exploits

    Solana Drainer Tool — Automated SOL and SPL token drainer for Solana wallets with real-time balance monitoring, multi-wallet sweeping, Phantom/Backpack/Solflare compatibility, and stealth transaction …

    pearlgriffin2021Python★40

  • D whatever The content is a one-line personal bio: 19 years old, a major, loves cheesecake, with no code or explanation; the twenty-odd stars only reflect social sharing and don't make it a project. #Social#Documents#Learning

    zahra, 19, hkbk, cse, 1st yr, strawberry cheesecake

    moonsklkLanguage unknown★28

Show HN 3

B

An analytics engine for Iceberg tables claiming to be 2x faster than ClickHouse/DuckDB, hitting the real pain point of query latency in data platforms where willingness to pay is high. But only 10 points in two days and no star evidence, the performance claim is unreproduced by third parties; the commercialization path is clear, so first see if it passes benchmark reproduction.

#Open-Source Alternatives#Data Analysis#clickhouse

▲10

Hot New Repos

  1. nullmoth/nvidia-macos-driver★1052 Metal driver for NVIDIA GeForce RTX cards on macOS 15 Sequoia (Intel / OpenCore). Free, source included.
  2. alejandrobujan/tendedero★991 Screenshots, hung out to dry. A tiny native macOS app that hangs every screenshot on a line at the top of your screen.
  3. LoreanXavier/pt-pc★980 Native PC port of P.T. (runs from your own PS4 game files)
  4. storytold/wordcraft★808 An open-source, clean-room reimplementation of Microsoft Word in pure Rust
  5. storytold/cadcraft★746 CADCraft: computer-aided design and drafting — an open-source, clean-room AutoCAD-style app in pure Rust
  6. shadany7824/playgta5★703 A source code for playgta5.com.
  7. mhtsec/ARTEX★696 AI autonomous penetration testing system | Champion project of Baidu's "agent+" attack-defense challenge
  8. pingdotgg/ts-rust★693 An experimental Rust port of the TypeScript 7 compiler (tsc)
  9. bas3line/ascii★597 Animated ascii art for web pages, in TypeScript: React, Next.js, Astro, or one HTML tag
  10. storytold/gridcraft★546 GridCraft: an open-source, clean-room spreadsheet (Microsoft Excel-style) in pure Rust. By ArtCraft.

Star gain 1d

  1. morluto/rea+12454★38713
  2. boykopovar/AnyPS5+10534★19410
  3. storytold/photocraft+9782★26715
  4. storytold/artcraft+7500★10151
  5. eternity4719/HowToLiveBetter+4228★54468
  6. mattpocock/skills+3247★282167
  7. cathrynlavery/diagram-design+2810★47411
  8. openai/math+2673★12059
  9. storytold/filmcraft+2000★5271
  10. anthropics/knowledge-work-plugins+1850★28008