Tools Bench.

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

Last Brew Time: Jun 23, 2026, 7:38 AM PT

Insight

Builders have stopped shipping apps for humans to click and started shipping tools that agents themselves operate

Featured

GitHub73.7K

An open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of tasks that could take minutes to hours.

Market Signal

Why It Has Market Pull

DeerFlow is ByteDance's open-source long-horizon SuperAgent harness that researches, codes and creates using sandboxes, persistent memory, skills, sub-agents and a message gateway. Built on LangGraph/LangChain, its v2.0 rewrite topped GitHub Trending and it has grown into one of the largest corporate-backed agent frameworks.

  • Roughly 73,773 stars and 9,961 forks, well above early press figures
  • Hit #1 on GitHub Trending within 24 hours of the v2.0 launch (Feb 27-28, 2026)
  • Backed by ByteDance; built on LangGraph/LangChain with Docker/Kubernetes sandboxes and MCP support
  • Actively developed: ~951 open issues with last push June 23, 2026 and weekly progress reports
  • Integrates Telegram, Slack, Feishu, WeChat and a Claude Code skill, plus LangSmith/Langfuse observability

feedbacks

What People Are Saying

  • "Its Arxiv search is particularly impressive, saving researchers a great deal of time."Dev.to article

  • "Multi-session continuity is a known gap and memory retrieval is inconsistent."Dev.to article

  • "FULLY FREE local deep research agent that can do almost anything."YouTube post

  • "ByteDance DeerFlow: deep research agents with a local LLM."YouTube post

  • "ByteDance's DeerFlow hits #1 on GitHub Trending."YouTube post

  • "Code execution tasks were reliable but not sophisticated."Dev.to article

GitHub32.9K

The open-source AI voice studio. Clone, dictate, create.

Market Signal

Why It Has Market Pull

Voicebox is an open-source, local-first AI voice studio that clones any voice, generates speech across 23 languages and seven TTS engines, powers system-wide dictation, and lets MCP-aware agents speak in a chosen voice, all without sending audio off-device. It is built by Jamie Pine, the developer behind the Spacedrive file manager, under an MIT license.

  • Roughly 32,968 stars and 3,982 forks (created January 25, 2026)
  • Created by Jamie Pine, the developer behind Spacedrive, lending strong author credibility
  • 25 releases; v0.5.0 (April 25, 2026) expanded it into full voice I/O with agent integration
  • Press coverage describes cloning from ~3 seconds of audio and running fully locally
  • Built on Tauri (Rust) + FastAPI with Qwen3-TTS and MLX/PyTorch inference, positioned as a free local ElevenLabs alternative

feedbacks

What People Are Saying

  • "Clones any voice from 3 seconds of audio and runs locally for free."TechTimes article

  • "The free local ElevenLabs alternative that just hit 22k GitHub stars."blog post

  • "Get started with Voicebox: an open-source alternative to ElevenLabs."YouTube post

  • "I tried the open-source ElevenLabs alternative."YouTube post

  • "MCP-aware agents can speak in cloned voices via a single tool call."GitHub README

  • "Hundreds of open issues suggest growing pains alongside heavy real-world usage."GitHub issues

HF Spaces16.6K 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.

Market Signal

Why It Has Market Pull

DeepSite is a free, open-source vibe-coding platform that turns a plain-language prompt into a full website or app with live preview and instant free hosting. With roughly 16,617 likes it is one of the most-liked Spaces on Hugging Face, runs in production, and has been iterated through v4 with a current 2026-era multi-model roster.

  • Roughly 16,617 likes, among the most-liked Hugging Face Spaces, with 566+ community discussion threads
  • Live in production (MIT, Docker); v3 announced Sept 25, 2025, now on v4
  • v4 multi-model lineup: DeepSeek-V3.2, Qwen3-Coder-30B, Kimi-K2-0905, GLM-4.7, MiniMax-M2.1
  • Built by a Hugging Face staffer and amplified by HF leadership and developer advocates
  • Original release went viral as a free web-based Cursor/Bolt alternative, spawning multiple clones

feedbacks

What People Are Saying

  • "Is it perfect? No. But is it dang amazing? Yes."blog post

  • "Building a site is as easy as microwaving popcorn."Medium article

  • "The distance between idea and execution just disappeared."Medium article

  • "Most sites come out static and complex prompts need manual editing, so keep it simple to start."Medium article

  • "DeepSite and DeepSeek are completely open-source; time for iterative improvements."Hugging Face community

  • "DeepSite vibe coding is going viral with awesome projects."Hugging Face community

  • "This free Deepseek AI coder is actually amazing."YouTube post

Market Signal

Why It Has Market Pull

Palmier Pro is a Swift-native macOS video editor built so AI agents (Claude, Codex, Cursor) can operate the timeline directly through a built-in MCP server, with in-timeline generative AI from models like Seedance, Kling, Veo and Nano Banana Pro. It is the open-source product of Palmier, Inc., a Y Combinator S24 company, free under GPLv3 with paid generative credits.

  • Roughly 8,279 stars and 541 forks (created April 7, 2026)
  • Backed by Palmier, Inc. (Y Combinator S24), confirmed via the YC company and Launch YC pages
  • Very active cadence: 55 releases, latest v0.4.0 on June 23, 2026
  • Founders report producing 15+ cinematic AI launch videos for YC companies with it
  • Requires macOS 26 on Apple Silicon, which limits the addressable audience today

feedbacks

What People Are Saying

  • "AI baked into the layer underneath the track, not bolted on as a sidebar panel."Hacker News

  • "Claude can now edit your video timeline; is 'Cursor for video' finally real?"YouTube post

  • "The free GitHub AI video editor that replaces paid alternatives."YouTube post

  • "I let AI edit a real video; here's what happened."YouTube post

  • "An open-source video editor your agents can operate."YC launch post

  • "Requires macOS 26 on Apple Silicon only, which limits who can run it today."GitHub README

Product Hunt177

The moment an agent needs to deploy something, it slams face-first into a wall built for humans. Today we're rolling out Temporary Accounts on Cloudflare Workers. Any agent can now run wrangler deploy — temporary and get a live Worker in seconds.

Market Signal

Why It Has Market Pull

A first-party Cloudflare platform feature, launched June 19, 2026 during Agents Week, that lets any AI agent run wrangler deploy with a temporary flag and get a live Worker in seconds with no signup. Backed by a major public company and bundled with Stripe and WorkOS agent-provisioning partnerships, it rides the existing Wrangler/Workers install base rather than needing standalone traction.

  • Official launch June 19, 2026 via the Cloudflare blog and developer changelog as part of Agents Week
  • Requires Wrangler 4.102.0+, distributed through the standard npm wrangler CLI
  • Shipped alongside a Stripe partnership and a WorkOS collaboration for agent account provisioning
  • Temporary accounts auto-expire after 60 minutes with proof-of-work and rate limits to curb abuse
  • Supports Workers, KV, D1, Durable Objects, Hyperdrive and Queues, but workers.dev only with no custom domains

feedbacks

What People Are Saying

  • "The AI hook isn't really necessary; this is an interesting feature for everyone else as well."Hacker News

  • "Amazing for PR previews and code review."Hacker News

  • "Wouldn't it make more sense to merge the temporary account into an existing one?"Hacker News

  • "I'm going to need a wrapper for all these services offering this."Hacker News

  • "Any agent can now run wrangler deploy and get a live Worker in seconds."Cloudflare blog

  • "Wrote a dedicated walkthrough after live-testing temporary Cloudflare accounts."Simon Willison blog

Sources

GitHub

220.3K

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

Use Garry Tan's exact Claude Code setup: 23 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA

Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface

LLM 驱动的多市场股票智能分析系统:多源行情、实时新闻、决策看板与自动推送,支持零成本定时运行。 LLM-powered multi-market stock analysis system with multi-source market data, real-time news, decision dashboard, automated notifications, and cost-free scheduled runs.

Product Hunt

Evaluate AI agents before they fail. Create test suites, run evaluations, and pinpoint issues before they reach production. AgentX provides full observability and traceability for your AI agents. AI analysis not only identifies problems but also suggests fixes-like an AI doctor for your agents. Simulate run your agents across multiple LLM providers to compare performance, cost, and latency, helping you make better decisions about which LLM to go. Run eval before deploy. Like CI/CD for AI agents.

MCP apps are the new website: they're real, interactive apps that run right inside AI assistants like ChatGPT and Claude. The tooling to build them didn't exist, so we built it. Meet Skybridge, the React framework that handles the infrastructure and the dev loop so you can focus on your features: MCP server, view rendering, client compatibility, testing tunnel, and more. Code once, ship everywhere. Already +500K downloads and dozens of apps in the stores!

We built HAQQ because understanding legal situations shouldn’t depend on who you know or whether you can afford to “just ask a lawyer.” Today, HAQQ Legal AI is available on mobile. Upload a contract. Ask a legal question. Get structured, jurisdiction-aware legal work with risk flags and exportable outputs. Not generic chat. Legal reasoning powered by Justinian®. Know exactly where you stand, wherever you are.

Alai captures every little detail about your brand in a design system. Use it to create presentations, social assets, ads, or any canvas size, all perfectly on-brand. Make precise edits manually or with AI. And choose your own AI models to balance cost, quality, and latency.

Your team’s work lives in project management tools and CRMs. Your exec work is scattered. readywhen catches every decision and commitment across Slack, email, meetings, and docs. Without being asked, it drafts what’s needed next: the brief, the email response, the update. You just approve. It's like having your own Chief of Staff, 24/7.

AirJelly is an always-on desktop AI agent that proactively captures your intent, organizes your tasks, and gets things done. By observing your on-screen activities, it knows who you are and what you're working on. It doesn't just spot your to-dos — it foresees what's next and acts. It also tracks your personal interactions, so you'll never miss a follow-up or forget a promise. Everything in your life is managed by AirJelly — freeing your time for what AI can't do, or simply for you to enjoy.

Hacker News

I'm the developer of Trace, a non-intrusive, shortcut-driven Mac app that records and transcribes your meetings on-device. I know, another meeting transcription app. Please bear with me though, I'm confident that this is at least a little novel. I primarily built Trace for myself. I'd been using MacWhisper, but there was enough fiddling before each call that I'd forget to start it and walk out of an hour-long meeting with nothing written down. So the things I cared about most were that it's quic... (205 points, 83 comments).

Anthropic and OpenAI's publicly available models are explicitly guard-railed so that they refuse offensive tasks. And their cyber-focussed models are gated for enterprises. This leaves SMEs and mid market open to major vulnerabilities. AI can be used as both an adversarial and defensive tool in the world of cyber. A worst case outcome is if only the adversaries have access. Meanwhile, most existing AI cyber tools are just wrappers. The problem is that they still have all the guardrails on from t... (92 points, 40 comments).

Hi HN, I'm Ahmad from the Intuned ( https://intunedhq.com ) team. Today, we're releasing and open-sourcing Selector Forge ( https://selectorforge.ai/ ), a browser extension that generates reliable CSS/XPath selectors using AI. You can use it to create a selector for a single element or for an array of elements. The selectors it creates are meant to be "semantic" and more resilient to page changes than what Chrome DevTool’s “Copy Selector” (and other similar extensions) give you. Those tend to ha... (34 points, 2 comments).

Shumai is an open source platform for uploading creative files, managing projects, collecting precise feedback, sharing work, and collaborating with AI agents, all in one simple creative-first workspace. I’ve always liked the product design of Frame.io, and I wanted to build an alternative that feels just as polished, while being open source and easy to self host. You can deploy Shumai with docker compose in just a few minutes. It can also be installed from npm, though you'll need to provide you... (24 points, 0 comments).

With AI, it's faster than ever to ship a marketing site... but most of what gets generated is slop that was never built to be found. Plus the tools meant to catch that fall short: most SEO auditors cost money, don't play nicely with your agents, or tell you what's wrong without telling you how to actually rank for SEO and GEO (Generative Engine Optimization: being cited by AI search like ChatGPT, Perplexity, and Google AI Overviews). crawlie fixes that! It's 100% free, it's local-first, it's age... (20 points, 6 comments).

Hey HN, meet Motion — a frontier agent for motion design. Here’s how it works: 1. Give Motion a prompt — include any context like product links, X threads, YouTube videos, personal assets, and design MD. Motion will reference styles, incorporate research, and storyboard scenes. 2. Motion creates the video — it can orchestrate entire explainers, launch videos, logo animations, convert articles to video, or even add motion design to your existing videos. 3. Edit everything — after generation, edit... (10 points, 3 comments).

HF Spaces

11.1K 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 is a Hugging Face Space tagged with gradio, region:us. It has 10113 likes on Hugging Face.

5.1K likes

Wan2.2 Animate is a Hugging Face Space tagged with gradio, region:us. It has 5113 likes on Hugging Face.

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...

Arena Leaderboard is a Hugging Face Space tagged with static, leaderboard, region:us. It has 4924 likes on Hugging Face.

The ultimate guide to training LLM on large GPU Clusters Instruction to install and run locally Loading HTML fragments: There are two way to load HTML fragments: Compile them into html during build time Fetch them and insert them during run-time When to use what Use compile time fragments only on parts which you want to ensure are seen by every user right after page load (e.g logo) Use run-time fragments for everything else so that the final HTML is of reasonable size (<1MB idealy) How to add a new fragment Add it to the src/fragments folder (e.g. src/fragments/banner.html) For run-time fragments, add {{{fragment-name}}} to appropriate place in src/index.html (e.g. {{{fragment-banner}}}) For...

AI Tools — June 23, 2026 Edition | Agentic Brew