AI agent skills: Why they succeed and what causes them to crash
Everyone's adding skills to their agents. Almost no one knows when they help, why they work, or where they fail.
// Intelligence
Today in AI
A focused feed of models, agents, research and open-source releases for people building with AI.
Updated
Pika API Club Lets Developers Access 100 AI Models for $10
Pika launches a $10/month API Club offering 100+ generative media models through one endpoint, at up to 88% below rival aggregator prices.
Liquid AI's LFM2.5 Beats Bigger Rivals in Real On-Device Phone Benchmarks
Artificial Analysis and Liquid AI launched a joint benchmark measuring how quantized small models actually perform on iPhone 17 Pro and Galaxy S26 Ultra.
ElevenLabs CLI v1 Moves Its Entire Voice Agent Platform Into the Terminal
ElevenLabs ships a terminal-first CLI that treats voice agents as version-controlled config files, with structured JSON and skills for coding agents.
Liquid AI's Pipette Exposes What Server Benchmarks Hide About Phone AI
Liquid AI and Artificial Analysis release an open-source suite that measures model quality, speed, latency, and memory across real phones, laptops, and embedded hardware.
CMU's Jeremy Avigad Says AI Math Is Bigger Than Theorem Provers
Jeremy Avigad argues the fixation on neural theorem provers hides a far richer landscape of ways AI is reshaping how mathematics gets done.
x64dbg-MCP Server Lets AI Agents Drive Windows Debuggers Autonomously
A new Zig-built plugin turns x64dbg into an MCP endpoint, letting Claude and other AI agents drive reverse engineering sessions through 71 tools.
Qwen Drops Qwen3.8-Flash-Next, a 125B Open Model That Activates Only 6B Parameters
Qwen ships an FP8 preview of the architecture behind Qwen4, pairing sparse attention, n-gram embeddings and 125B params with 6B active.
SparkLLM's Spark-X2.5 Packs 1M-Token Context Into a Tiny 4B Model
A compact 4B open-source model with hybrid sliding-window attention, native 1M-token context, and agent-focused benchmarks that top comparable small models.
Alibaba's CommerceAgentBench Reveals Top AI Agents Fail 38% of Real Business Tasks
Alibaba's Accio team open sourced a 107-task benchmark that forces agents to complete real e-commerce workflows inside stateful replicas of Shopify, Gmail, Stripe and more.
Alibaba's MiniMax-H3-Fun-Controlnet-Union Steers Video Generation With One 6.8 GB Checkpoint
Alibaba PAI drops a single ControlNet-Union checkpoint that adds Canny, Depth, HED, MLSD, Pose and inpainting to MiniMax-H3 video generation.
Someone Reverse-Engineered Grok Bot 0.18 and Rebuilt It With Four AI Backends
A weekend reverse-engineering project turns a compiled macOS app into 440K lines of readable TypeScript, then bolts on a four-provider inference router.
GitHub's Dependabot Now Waits 72 Hours Before Updating Your Dependencies
GitHub added a three-day default cooldown on Dependabot version update pull requests, aiming to filter out short-lived poisoned package releases before they land in your repo.
GooeyPi Unifies Pi, OMP, and Prime Agent Into One Open-Source Desktop
GooeyPi wraps three terminal coding agents in a single Electron desktop app, giving Pi, OMP, and Prime Agent a shared cross platform GUI.
Qwen3.8 Hits 49 Tokens per Second on AMD's Budget APU
A community quantization pairs a 27B Qwen model with multi-token prediction and AMD's IU4 matrix path, hitting ~49 tokens per second on a single Strix Halo APU.
Solo Creator Grafts Z-Image Textures Into MiniMax-H3 Without Any Training
A community-made checkpoint grafts Z-Image's texture attention onto MiniMax H3's video engine, giving richer surfaces without retraining or extra VRAM.
Anthropic's Claude Code Remote Control Finally Stays Connected on Mobile
Anthropic ships reliability fixes to Claude Code Remote Control, adds auto-reconnect, tighter phone-CLI sync, and the ability to launch sessions from your phone.
OpenAI Gives Developers Hard Spend Limits to Stop Runaway API Bills
OpenAI now lets teams attribute spend down to individual API keys and enforce hard monthly caps that cut off traffic when hit.
DeepSeek V4 Pro Hits 90% on ARC-AGI-1 but Stalls on Harder Puzzles
DeepSeek V4 Pro hits 90.5% on ARC-AGI-1 and 61.3% on ARC-AGI-2, but extra reasoning barely moves the needle on abstract puzzles.
Google Research's ME-POIs Beats Gemini Embeddings at Reading Real Places
Google Research unveils ME-POIs, a framework that fuses anonymized foot-traffic patterns with text embeddings to improve place understanding by up to 81.9%.
Anthropic Opens Claude Mythos 5 to Enterprise Teams for Hunting Code Vulnerabilities
Anthropic opens its most capable cybersecurity model to Enterprise customers through scans, partner integrations, and $35M in open-source credits.
Google's Biomarker Discovery Framework Finds 66 Health Signals Wearables Always Missed
Google's new multi-agent system turns messy wearable sensor streams into statistically vetted biomarker candidates through adversarial validation and human review.
Pika Labs' Pika Speech Undercuts ElevenLabs by 9x With Studio-Quality Audio
Pika Speech is a 3B flow-matching TTS model that generates a minute of 48 kHz audio in about a second at one cent per minute.
Codex Router Merges DeepSeek and Kimi Into OpenAI's Picker at 89x Lower Cost
A local router merges Kimi, DeepSeek, and other external models into the Codex picker without kicking out your native GPT models or ChatGPT login.
No stories match these filters.