Tools Bench.

Product launches and open-source repos with enough signal to earn a second look.

Last Brew Time: Jul 13, 2026, 10:26 AM PT

Insight

Builders are racing to wrap the same three coding agents rather than replace them

01.INSTALLABLE SKILLS

02.PERSISTENT AGENT MEMORY

03.MCP-CONNECTED TOOL ENDPOINTS

04.PROMPT-TO-APP VIBE CODING

Featured

GitHub120.4K

github/spec-kit

💫 Toolkit to help you get started with Spec-Driven Development

youtube
The ONLY guide you'll need for GitHub Spec Kit

Market Signal

Why It Has Market Pull

A GitHub-official, fast-moving open-source toolkit for spec-driven development that has become one of the most-starred AI engineering projects of the year, with a large contributor base and near-daily releases.

  • Roughly 120,000 stars and 10,700 forks since first published in August 2025
  • Around 239 contributors and 611 watchers - broad, distributed maintenance
  • Shipped v0.12.13 on July 13, 2026, with multiple point releases in the same week
  • Backed and hosted by GitHub itself under an MIT license, with an official docs site
  • Independent tutorial coverage on YouTube exceeding half a million views on a single explainer

feedbacks

What People Are Saying

  • "I'm really enjoying working with SpecKit"HN comment

  • "The constitution created as the first step is valuable in its own right"HN comment

  • "spec-kit is very helpful for both green/brown field projects"HN comment

  • "the tutorials are all still pretty elementary"HN comment

  • "a sea of markdown documents, long agent run-times and unexpected friction"Dev blog

  • "The overhead isn't worth it for small features"Dev.to article

  • "Spec-Kit shines - the agent thinks through edge cases you might have missed"Dev.to article

GitHub84.3K

Graphify-Labs/graphify

AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI, and more). Turn any folder of code, SQL schemas, R scripts, shell scripts, docs, papers, images, or videos into a queryable knowledge graph. App code + database schema + infrastructure in one graph.

youtube
The Truth About Graphify's 70x Token Saving Claim

Market Signal

Why It Has Market Pull

A genuinely viral open-source skill that turns a codebase or document folder into a queryable knowledge graph for AI coding assistants; the traction is real and single-repo, though its headline token-saving claim is heavily conditional and not yet independently reproduced at scale.

  • About 84,000 stars and 8,300 forks despite the repo only being created in April 2026
  • Roughly 141 contributors and near-daily releases, reaching v0.9.14 on July 13, 2026
  • Installs as a skill across 17 assistants including Claude Code, Cursor, Codex and Gemini CLI
  • Commercial homepage with a free local tool plus a paid enterprise platform
  • The widely cited ~71x token reduction is a best-case tail; one independent benchmark measured 7.3x

feedbacks

What People Are Saying

  • "Turns your codebase into a knowledge graph your AI queries instead of grepping"Dev.to article

  • "The 71.5x number is real but it's the tail of the distribution, not the middle"Medium review

  • "Under 100 files, Graphify barely helps"Medium review

  • "One independent benchmark achieved 7.3x, not the viral 71.5x"Exchangepedia review

  • "Graphify is an exploration and orientation tool, not an implementation tool"Medium review

  • "Does close to nothing for the tokens you spend changing code"Medium review

  • "84k stars on a repo created in April 2026 feels inflated"X reply

GitHub4.9K

Nutlope/hallmark

Anti-AI-slop design skill for Claude Code, Cursor, and Codex.

youtube
Fixing slop with Hallmark skills

Market Signal

Why It Has Market Pull

Hallmark is an open-source, anti-AI-slop design skill for Claude Code, Cursor and Codex from a marquee open-source builder, maintained with Together AI. It encodes typography, color, layout and motion rules plus a large slop-test gate to stop AI coding agents producing generic-looking sites, and it has grown fast since launch.

  • About 4,960 stars and 270 forks, created April 27, 2026
  • Installed 1,000+ times in the first 24 hours per the maker
  • Hit GitHub Trending (#17 on July 12, 2026) and shipped a v1.1 release with new themes and more slop detectors
  • 22 themes and four commands (build, audit, redesign, study) with a 65-gate slop test and pre-emit self-critique; MIT-licensed with a live showcase site
  • Distributed via a one-line skills install plus native Claude Code, Cursor and Codex paths

feedbacks

What People Are Saying

  • "Hallmark has already been installed over 1,000 times in 24 hours! Blown away by all the interest."X post

  • "Hallmark v1.1 is out - now with new themes, better designs, more AI slop detectors, and a new mode to redesign apps."X post

  • "A design skill that refuses to look AI-generated."usehallmark.com

  • "Two pages for different briefs should not share the same hero, 3-feature, CTA, footer rhythm."GitHub README

  • "Fixing slop with Hallmark skills."YouTube post

  • "A complete guide to creating non-AI-looking sites with Claude Code using Hallmark."Dev walkthrough

  • "Featured on GitHub Trending, first reaching #17 on Jul 12, 2026."Trendshift

HF Spaces3.5K likes

Z Image Turbo

Z Image Turbo is a Hugging Face Space tagged with gradio, mcp-server, region:us. It has 3543 likes on Hugging Face.

youtube
Ultimate Z Image Turbo Guide + ComfyUI (Install, LoRA, Inpaint, Img2Img)

Market Signal

Why It Has Market Pull

A fast open-source text-to-image demo wrapping Alibaba Tongyi's 6B Z-Image-Turbo model, exposed as a live Gradio Space with an MCP server. The underlying model has crossed one million downloads and the Space has thousands of likes, with a large creator ecosystem building workflows around it.

  • About 3,545 likes on the Hugging Face Space (Gradio, live), tagged mcp-server
  • Parent model from Alibaba Tongyi has about 4,977 likes and over 1,011,000 downloads
  • 6B-parameter model generating images in under ~5 seconds (~3s on an RTX 4090 per user reports)
  • Deep third-party tutorial ecosystem: ComfyUI guides at 22K-44K views covering ControlNet, LoRA, inpaint and upscaling
  • Open-source with a permissive posture, positioned as an alternative to Flux and Midjourney

feedbacks

What People Are Saying

  • "It's fast (~3 seconds on my RTX 4090)"HN comment

  • "for a 6B model that can generate an image in under 5 seconds, it punches far above its weight class"HN comment

  • "The adherence is impressive for a 6B parameter model"HN comment

  • "Surprisingly capable of maintaining image integrity even at high resolutions"HN comment

  • "This Open-Source AI Changed Image Generation Forever!"YouTube post

  • "You Won't Believe How Fast Z-Image Turbo Runs on Low-End PCs"YouTube post

  • "Union 2.1 - Better Inpaint & Control for Z-Image-Turbo in ComfyUI!"YouTube post

Product Hunt413

FetchSandbox

Most API tests stop at 200 OK. FetchSandbox lets developers and AI agents verify what happens next—webhooks, retries, state changes, async workflows, and failure scenarios. It reproduces the real bug, proves the fix, and remembers what breaks—so your agent catches it before production. Connect via MCP to Cursor, Claude Code, Windsurf, VS Code, and Codex. Explore 60+ APIs—Stripe, GitHub, Clerk, Resend, Twilio, Descope, OpenAI—without burning real API quota or waiting on staging.

youtube
FetchSandbox MCP: Run API workflows from Cursor, Claude, or your IDE

Market Signal

Why It Has Market Pull

FetchSandbox is a stateful API-integration sandbox that lets developers and AI agents verify webhooks, retries, state changes and failure scenarios against 60+ real API specs without burning live quota, wired into IDEs over MCP. It launched to a strong reception and fast early adoption.

  • Reached #3 Product of the Day on Product Hunt with roughly 414 upvotes shortly after its July 12, 2026 launch
  • Crossed 1,600+ MCP installs across Cursor, Claude Code, Windsurf, VS Code and Codex
  • Ships 60+ API specs including Stripe, GitHub, Clerk, Resend, Twilio and OpenAI, plus a published npm package and a Glama MCP-server listing
  • Distributes an official demo video and a developer write-up explaining the stateful-sandbox approach
  • 4.3 rating across early reviews

feedbacks

What People Are Saying

  • "the failure-library angle is interesting but the thing I'd want proven: who keeps the simulated failure behavior honest against the real API over time?"Product Hunt comment

  • "a fully automated per-api drift monitor that continuously diffs our simulation against the live api isn't there yet."Product Hunt comment

  • "The mock told me POST /charges returns an id. But my code then called GET /charges to verify status, and the mock returned 404."Dev.to article

  • "POST creates a real resource I can GET later; IDs chain between requests like they would in production."Dev.to article

  • "test API integrations your AI agent wrote (50+ specs, 30s install)."Cursor community forum

  • "verify what happens next - webhooks, retries, state changes, async workflows - not just a 200 OK."Product Hunt comment

  • "The MCP server exposes three tools: ingest an OpenAPI spec, list its workflows, and run them with schema-validated responses."Glama MCP registry

Sources

GitHub

moeru-ai/airi

41.8K

💖🧸 Self hosted, you-owned Grok Companion, a container of souls of waifu, cyber livings to bring them into our worlds, wishing to achieve Neuro-sama's altitude. Capable of realtime voice chat, Minecraft, Factorio playing. Web / macOS / Windows supported.

Raphire/Win11Debloat

50.7K

A simple, lightweight PowerShell script that allows you to remove pre-installed apps, disable telemetry, as well as perform various other changes to declutter and customize your Windows experience. Win11Debloat works for both Windows 10 and Windows 11.

Product Hunt

Miora

578

Not another AI image tool. Miora is an Agentic Creative Studio with Memory. Bring one idea, generate multimodal assets on one editable canvas, turn auto-built memory into a reusable Skill, then create more, always true to your taste. One person, a whole creative studio.

JustVibe

563

JustVibe is a free search engine, built to help get things done. Search "plan my 5 day trip to tokyo" and instantly receive a fully functional and interactive trip planner, perfectly set up for Tokyo and running directly in your browser. If your perfect app doesn't exist yet, JustVibe builds a custom one for your exact needs in minutes. Every app is yours to keep forever, with zero code. Chat to customize every detail and share your new app with a single link.

Second Brain for AI v2

365

Second Brain remembers your projects, people, decisions, and preferences across Claude, ChatGPT, Cursor, Codex, and any MCP client. V2 automatically links related memories, follows those connections during recall, and distinguishes settled decisions from drafts and stale context. Open source and self-hosted in your Cloudflare account.

ServiceBeard

159

Turn your customer-facing mailbox into an issue board. ServiceBeard syncs your emails directly to your issue tracker, letting you run a full service desk from your existing workspace. You can leverage the automation pipelines you’ve already set up without investing in expensive, per-seat helpdesk software. It's open-source, connects via standard IMAP/SMTP, and currently supports GitHub, GitLab, and Linear.

Offer Max

33

Finding a job is already a full-time job. Offer Max makes it a lot less painful. The Chrome extension captures any Job Description in one click. AI tells you how good your fit is before you commit to applying, rewrites your résumé in the recruiter's language, drafts the cover letter, autofills the tedious application forms, and preps you with the interview questions this team is likely to ask. Build your master profile once. Show up more prepared than everyone else.

CONTRABAND VR Cinema

32

Put on your Quest headset and step into a real cinema with a real audience, watching the same film at the same moment together. Nobody pauses and nobody skips ahead. When the credits roll, you vote on what screens next. This is what going to the movies used to feel like. This is Virtual Cinema.

YC Launch

Petrarch — The market maker for data

4

The market for specialized data locked up in enterprise companies. Petrarch · Summer 2026 · B2B Tags: Marketplace, B2B, Data Engineering. Website: https://petrarch.co/

Assemble: Agents for Enterprise IT

3

Agents that perform IT implementations across ERPs, CRMs, and HRIS platforms. Assemble · Summer 2026 · B2B Tags: SaaS, Enterprise Software, AI. Website: https://www.assemble.ai/

PathPilot - AI Platform for Lenders

2

AI agents that help lenders grow without scaling operational headcount. PathPilot · Summer 2024 · B2B Tags: Finance, B2B, Lending, AI. Website: https://www.getpathpilot.com/

Rise Reforming: distributed chemical production

We turn waste gases into valuable supply-secure chemicals. Rise Reforming · Summer 2026 · Industrials Tags: Hard Tech, Smart Waste Management, Alternative Fuels, Renewable Energy, ClimateTech. Website: https://rise-reforming.com/home

Hacker News

Reverse-engineering web apps into agent tools

96 pts

Hey HN! We built a browser-based agent that runs inside an authenticated web app, watches how the app calls its own APIs, and automatically turns those into agent tools. You can think of it as an auto-generated MCP server that self-updates as the host app changes. The result is a skilled AI assistant that actually integrates deeply with any product (not just chat and RAG) with minimal effort. Check out these short demos below that show the agent in software you're probably familiar with: - Jira:... (96 points, 40 comments).

I built a web tool to see and edit what an AI thinks before it answers

34 pts

I run a small AI lab and playground and got super excited about Anthropics paper "Verbalizable Representations Form a Global Workspace in Language Models" ( https://transformer-circuits.pub/2026/workspace/index.html ) It talks about how they use a tool they call a Jacobian Lens to view inside the middle layers of LLM while it's working before it commits to a word (token). I wanted to see if I could get a version of this running on the open models and to my surprise it worked! I ran some experime... (34 points, 8 comments).

AI harness for C/C++ with GDB, sanitizers, perf and compile tools

5 pts

hey guys, i wanted to show one of my side projects. The idea is a coding harness (independent of models) natively designed for C/C++ developer workflows. I'm a C++ dev and do not find claude code work well with C++ toolchain like gdb and perf. The current version has integrations for gdb, clang-tidy, cppcheck, sanitizers, perf, benchmarking, compile DB navigation, Godbolt, symbolization, binary inspection, and decompilation. It supports Anthropic, OpenAI, Gemini, and self-hosted models. There ar... (5 points, 2 comments).

Clx – Compile Lua to Native Executables Through C++20

65 pts

Hi HN, clx is an ahead-of-time compiler for standard Lua that generates C++20 and produces standalone native executables through GCC, Clang or MSVC. The project started as an experiment to see whether modern C++ could be used as a portable compiler backend instead of LLVM or direct machine code generation. The generated code is then compiled and optimized by the host toolchain. The latest release replaces the previous NaN-tagged value representation with a new shadow-types implementation, adds f... (65 points, 0 comments).

I hated how much my 12-year-old played Roblox, so we built our own FPS

23 pts

I'm a father of two, 7 and 12. They are obsessed with Roblox, especially Rivals. Like a lot of parents, we did not love it. We tried the usual things: block it, limit it, set timers." It became a daily battle, a lose-lose situation. So I flipped the problem. Instead of fighting what they loved, I decided to lean into it, but with a twist. Why just play an FPS when you could build one together? My kids became the PMs. Claude and I became their engineer. I was shocked by how fast we moved. We pick... (23 points, 7 comments).

HF Spaces

DeepSite v4

16.6K likes

Generate any application by Vibe Coding it DeepSite is a Vibe Coding Platform designed to make coding smarter and more efficient. Tailored for developers, data scientists, and AI engineers, it integrates generative AI into your coding projects to enhance creativity and productivity. DeepSite v4 is a Hugging Face Space tagged with docker, region:us. It has 16617 likes on Hugging Face.

Open LLM Leaderboard

14.0K likes

Track, rank and evaluate open LLMs and chatbots Modern React interface for comparing Large Language Models (LLMs) in an open and reproducible way. 📊 Interactive table with advanced sorting and filtering 🔍 Semantic model search 📌 Pin models for comparison 📱 Responsive and modern interface 🎨 Dark/Light mode ⚡️ Optimized performance with virtualization The project is split into two main parts: React Material-UI TanStack Table & Virtual Express.js FastAPI Hugging Face API Docker The application is containerized using Docker and can be run using: Open LLM Leaderboard is a Hugging Face Space tagged with docker, leaderboard, modality:text, submission:automatic, test:public. It has 14039 likes on Hu...

AI Comic Factory

11.2K likes

Create your own AI comic with a single prompt Last release: AI Comic Factory 1.2 The AI Comic Factory has an official website: aicomicfactory.app For more information about my other projects please check linktr.ee/FLNGR. If you like the AI Comic Factory, let me know! I am always creating new spaces and exploring new ideas for demos, meaning I don't have much time to take care of all of them (I wish I could clone myself or ask robots to do it). If you appreciate the AI Comic Factory and would like to leave a tip, that would be very kind 🫶 First, I would like to highlight that everything is open-source (see here, here, here, here). However the project isn't a monolithic Space that can be dupli...

Kolors Virtual Try-On

10.1K likes

Kolors Virtual Try-On is a Hugging Face Space tagged with gradio, region:us. It has 10129 likes on Hugging Face.

Wan2.2 Animate

5.1K likes

Wan2.2 Animate is a Hugging Face Space tagged with gradio, region:us. It has 5116 likes on Hugging Face.

MusicGen

5.1K likes

AudioCraft is a PyTorch library for deep learning research on audio generation. AudioCraft contains inference and training code for two state-of-the-art AI generative models producing high-quality audio: AudioGen and MusicGen. Installation AudioCraft requires Python 3.9, PyTorch 2.0.0. To install AudioCraft, you can run the following: We also recommend having ffmpeg installed, either through your system or Anaconda: At the moment, AudioCraft contains the training code and inference code for: MusicGen: A state-of-the-art controllable text-to-music model. AudioGen: A state-of-the-art text-to-sound model. EnCodec: A state-of-the-art high fidelity neural audio codec. Multi Band Diffusion: An EnC...

AI Tools — July 13, 2026 Edition | Agentic Brew