Tools Bench.

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

Last Brew Time: Aug 5, 2026, 10:43 AM PT

Insight

The loudest pattern in this run isn't a new AI category at all: it's how much of the pile is duplicates and near-zero-traction copies dressed up as fresh launches

01.ANONYMOUS RE-HOSTS OF THE SAME OPEN MODEL ON HUGGING FACE

02.SAME-DAY SHOW HN LAUNCHES NOBODY ENGAGED WITH

03.THE RARE REAL MOMENTUM INSTEAD CONCENTRATED IN BIG-LAB INFRASTRUCTURE DROPS

Featured

GitHub14.9K

TencentCloud/TencentDB-Agent-Memory

TencentDB Agent Memory is a team-level memory hub for AI Agents — turning conversations, docs, and code into four reusable memory assets (Chat Memory, Skill, LLM-Wiki, Code-Graph) that are governed, shared, and equipped across agents and frameworks.

youtube
China Just Open-Sourced Humanlike Memory for AI Agents (Tencent DB)

Market Signal

Why It Has Market Pull

A team-level shared memory layer for AI agents from a major cloud provider that is currently trending #1 on GitHub, gaining over a thousand stars a day. Independent testers have already reproduced its core claims on consumer hardware and published substantive technical critiques — unusually fast, credible validation.

  • 14,900+ stars and 1,360+ forks within about four months, trending #1 on GitHub
  • Backed by a major cloud provider, giving enterprise distribution and credibility
  • Shipped a v2.0 release with integration hooks for multiple agent frameworks
  • Independently reproduced by a third-party reviewer on a local model on a laptop
  • 500+ open issues and hundreds of open PRs signal heavy real usage

feedbacks

What People Are Saying

  • "the most complete open team memory control plane we have seen in this wave"Independent blog review

  • "A memory hub stores; it doesn't adjudicate. That distinction is the whole ballgame once memory becomes shared"Dev.to article

  • "confident recall of stale facts is worse than no memory"Dev.to article

  • "by turn 20 your context is 80% archaeological log dust the model is paying attention tax on every step"Medium article

  • "it fires at edge scale"Medium hands-on test

  • "Treat v2 as production-curious: star count and PR volume signal heat and unfinished edges"Independent blog review

GitHub11.2K

firecrawl/pdf-inspector

Fast Rust library for PDF inspection, classification, and text extraction. Intelligently detects scanned vs text-based PDFs to enable smart routing decisions.

youtube
節省 54% OCR 成本！pdf-inspector 實戰：打造高效能文件自動化流程

Market Signal

Why It Has Market Pull

pdf-inspector is a genuine, high-momentum open-source release from Firecrawl, announced via official company blog and founder posts as the engine behind their new PDF parsing pipeline. It hit #1 on GitHub Trending with explosive star velocity and substantive ongoing engineering activity.

  • 11,280 stars / 748 forks, reached #1 on GitHub Trending within days of release
  • Independent star-tracking service measured ~1,717 stars/day, calling it top 0.1% of tracked repos
  • Announced by Firecrawl's official blog and via posts from the company and founder accounts
  • 271 GitHub issues with substantive engineering discussion (regression corpora, eval-diff review, OCR routing)
  • Claims 3.5-5.7x speed improvement over Firecrawl's prior parser, used in production

feedbacks

What People Are Saying

  • "we open sourced the fastest pdf parser engine, powers /parse together with our custom OCR models"X (company account)

  • "we built pdf-inspector so agents can process PDFs without expensive OCR calls"X reply (founder)

  • "near the top of the star-velocity board... top 0.1% of the roughly 3,300 repos I track"GitHub issue comment

  • "Holding this — reviewing the eval diffs turned up two failure modes"GitHub issue

  • "Ruled table splits into multiple tables when a single cell wraps to more than one line"GitHub issue

  • "Corpus net went down; I stated the effect backwards"GitHub issue comment

GitHub2.3K

cloudflare/computer

Give your agent a computer 👾

Market Signal

Why It Has Market Pull

cloudflare/computer is Cloudflare's open-source "give your agent a computer" project — a virtual filesystem and execution runtime for AI agents, released as part of Cloudflare's broader Agents Week push. It's a credible, well-resourced company entering the agent-infrastructure space with real engineering depth and front-page developer discussion.

  • 2,549 GitHub stars and 116 forks in about two months, with same-day commits
  • Related Hacker News thread drew 151 points and 72 comments
  • Companion Cloudflare blog post ties it to Cloudflare's paid Sandboxes/Workers platform
  • Built and championed publicly by Kenton Varda (creator of Cap'n Proto / Sandstorm.io)
  • Explicitly labeled a preview with unstable APIs — not yet production-ready

feedbacks

What People Are Saying

  • "pandoc --pdf-engine fails with "openFile: resource busy (file is locked)" on deployed Containers"GitHub issue

  • "`edit` tool rewrites the whole file when `oldText` doesn't match exactly"GitHub issue

  • "he already built this once, a decade ago, and it failed for a reason AI may have just removed"HN comment

  • "don't trust your agent to not access your files"HN comment

  • "Add a write capability for a single command, plus gate and audit seams"GitHub issue

  • "suitable for experiments, exploration and prototypes but not suitable for production use at this time"GitHub README

Sources

GitHub

obra/superpowers

267.2K

An agentic skills framework & software development methodology that works.

esengine/DeepSeek-Reasonix

31.5K

DeepSeek-native AI coding agent for your terminal. Engineered around prefix-cache stability — leave it running.

huangruiteng/loopx

2.0K

Lightweight loop engineering state kernel for long-running AI agent teams. Agent-loop agnostic across Codex, Claude Code, and other coding agents, with durable goals, quota-aware auto-wake, executable todos, evidence logs, and verifiable handoffs.

Product Hunt

Hey Noah

559

Noah is a proactive AI executive assistant for founders. Claude talks to you; Noah talks to your network. It's like Tesla Full Self Driving, and not cruise control- an autonomous AI EA that manages your calendar, relationships, and follow-ups across email, text, and WhatsApp - so nothing slips. Think of it as an always-on EA living in your SMS and helping you with meeting logistics.

Wondering

367

Wondering is the most delightful way to break down complex topics into knowledge you can remember and apply. Tell it what you want to learn, and it creates a personalized path of short lessons with visuals, podcasts, and interactive exercises. It surfaces the most important ideas and lets you explore them in your own way, like a thoughtful tutor beside you.

Driven

265

Driven is an AI investment agent that turns market insight into action, not just answers. In one agent-powered workspace, it combines 260+ APIs, built-in & custom Skills, Playbooks, scheduled tasks, 24/7 monitoring, real-time data, portfolios, and order workflows. Instead of switching tools or chasing every signal, investors can manage the full journey in Driven, from idea generation and data gathering to analysis and action. It’s your AI investment team, while you stay in control.

Atlaso

250

Atlaso is a memory layer for AI. Connect it once and every AI you use, from Claude Code to Cursor, Codex and ChatGPT, automatically recalls the context that matters: your projects, your decisions, and the way you like to work. No more re-explaining yourself at the start of every session. One shared memory that follows you across every tool, instead of being locked inside one app. Free to start, and backed by original memory research.

Vibe Buddy

172

Vibe Buddy is a small desktop robot that keeps Codex and Claude Code usage limits and task status visible while you work. See when your AI is working, needs your input, has finished, or has hit its usage limit—without opening another tab.

ZapDigits MCP

134

Connect Claude and ChatGPT to Google Analytics, Search Console, Meta Ads and 30+ marketing data sources.

YC Launch

Dailybot CLI: Your AI agents now report to the team

10

An open-source CLI and skill that turns private agent work into team updates, approvals, and workflows in chat Dailybot · Summer 2021 · B2B Tags: Artificial Intelligence, Developer Tools, SaaS, Collaboration, AI Assistant. Website: https://www.dailybot.com/

Instaplay – Make and play anything

2

Every game imaginable is about to exist, and you'll find it or make it on Instaplay. Make and play anything in minutes, then share it with friends. Instaplay · Summer 2026 · Consumer Tags: Gaming, AI. Website: https://instaplay.ai

Hacker News

Draco – A single-binary, self-hostable Firecrawl alternative in Rust

14 pts

Scraping modern websites has become a massive headache. You basically have two choices: pay for an expensive API like Firecrawl/Browserbase, or run a fleet of headless Chrome instances that eat 1GB of RAM per page and still get blocked by Cloudflare. I built Draco to fix this. It’s a fast, single-binary web scraper written in Rust. You point it at a URL, and it spits out perfectly clean Markdown or structured JSON for LLMs. The secret sauce is that it doesn't just boot a browser for every reques... (14 points, 10 comments).

Blender for AI Agents

6 pts

Creating 3D is hard. LLMs seem to be getting better at tool use and spatial understanding. While MCPs have proved to be a good way to use these tools- the current methods have these challenges: - Access to scene graph and core C modules of Blender - Lack of parallelism, only way is to run blender headless - Lack of deterministic and fast verification layer - Inference stack- only way to use inference is to hook another MCP We're building Mixar, think Cursor for 3D. One access point to all genera... (6 points, 6 comments).

Sprocket – The Best AI Agent for Hardware and Software Development

124 pts

Hey HN, I am 16y/o and have been working on Sprocket for a while. It's an open-source AI agent that beats every other agent out there at both hardware and software. And here's the best part: Sprocket can (on its own) buy anything from any website when you tell it to do so. From hardware parts to SaaS subscriptions. Sprocket retrieves best-in-class context from the web for everything it does. It is therefore incredibly reliable. The agent harness's quality, performance, and UI rival that of Codex... (124 points, 14 comments).

Wienerdog – memory and self-improving skills for Claude Code/Codex

9 pts

The idea of Wienerdog was born out of my experience setting up my own simple but effective system of memory, self-improving skills and hooks for Claude Code and Codex. As I was teaching my friends and colleagues how to set up their own I found myself automating more and more of my system setup and finally I decided to publish it on GitHub to help others get more out of their AI usage. So what is Wienerdog? Simply put, it is just files — no daemon, no server, no telemetry: a collection of instruc... (9 points, 2 comments).

HF Spaces

Qwen-Image-Edit-2511-LoRAs-Fast

2.3K likes

Demo of the Collection of Qwen Image Edit LoRAs Qwen-Image-Edit-2511-LoRAs-Fast is a Hugging Face Space tagged with gradio, mcp-server, region:us. It has 2266 likes on Hugging Face.

FLUX.2 Klein multi-LoRA

321 likes

Use multiple FLUX.2-Klein LoRAs Note: This space is experimental and may log image-uploads during certain periods for performance monitoring. Always comply with HF and model Terms of Service. Any stored images are automatically deleted after 7 days. FLUX.2 Klein multi-LoRA is a Hugging Face Space tagged with gradio, mcp-server, region:us. It has 321 likes on Hugging Face.

Inflect v2 — Tiny Local TTS

112 likes

Run complete 3.96M and 9.36M text-to-waveform models live. Live text-to-waveform inference for both Inflect v2 release models: Inflect-Micro-v2: 9.36M parameters Inflect-Nano-v2: 3.96M parameters Choose the runtime that fits your device: ZeroGPU: server-side generation in this Space, with no local model download. Browser WebGPU: private, queue-free on-device inference in the same interface, with optional streaming and a WASM compatibility fallback. Every result is synthesized live from text. There is no reference audio, prerecorded fallback, or inference-time teacher model. Use the Compare tab to run the same text, speed, variation, and seed through both checkpoints. Long input is split auto...

Mage-VL

98 likes

Codec-native video & image understanding with Mage-VL 4B Mage-VL — codec-native streaming multimodal model Demo of microsoft/Mage-VL, a 4B codec-native vision-language model (Mage-ViT encoder trained from scratch + Qwen3-4B-Instruct-2507 decoder). Instead of decoding video into uniformly sampled frames and pushing a dense grid of patch tokens through a ViT, Mage-VL follows the structure of a video codec: it keeps every anchor (I) frame patch and only the predicted (P) frame patches where the codec spends bits — the regions carrying real motion and new detail. Those surviving patches are packed into canvases, cutting visual tokens by >75%. Image — single-image Q&A. Video — video Q&A, switchab...

MiniMax H3

95 likes

Video generation with a synchronized soundtrack MiniMax-H3 — unquantized, split across two Spaces Joint video and soundtrack out of a single denoising pass, at bfloat16 with no quantization anywhere. This Space is the denoising half: the 61.73 GiB transformer and the two autoencoders. The 62.14 GiB Qwen3-VL conditioner runs in qwen3vl-conditioner, which this Space calls over the gradio API for every request. The weights are the public MiniMaxAI/MiniMax-H3 diffusers checkpoint. MiniMax-H3 is 195.9 GiB in bfloat16 and a ZeroGPU Space is evicted at 150 GB of storage. An unquantized single Space is therefore impossible, which is why quantized demos of it run NVFP4 or float8 weights. Cut the Mini...

ABot-World Interactive

88 likes

Live interactive world rollout from an image 🌍 ABot-World — Interactive World Rollout Upload a single starting image and steer a live navigable world in real time. Provide a first-frame image (image-to-video seed), describe the scene, and drive the world with WASD (move & turn) / IJKL (look & pan). The model autoregressively rolls out an action-conditioned world and streams decoded frames straight to your browser. Model: acvlab/ABot-World-0-5B-LF (built on Wan2.2-TI2V-5B) Code: amap-cvlab/ABot-World Project: ABot-World This Space runs a proper live backend/infrastructure (in the spirit of Overworld/waypoint-1-5): gradio.Server exposes ZeroGPU-friendly /startgame and /stopgame API endpoints p...