Testing an API used to be simple. Then agents showed up, and you don't know which one your user is running or what it decides to do on its own.
AXIS runs real agents against your service and scores the experience.
Free and open source: axis.run 
Fable 5 vs GPT 5.5 vs GLM 5.2 vs DeepSeek V4 Pro
We tested 4 models in one game. Each result was one-shotted with the same precise prompt to compare capabilities.
For the OS model, we used /𝒅𝒆𝒔𝒊𝒈𝒏 from Command Code.
We’re really surprised that OS models can surpass top-tier models at this level. 
❤️ epanet-js is the one piece of software where they announce a 'pipe library' and it's not the pipe operator or visual programming but a literal library of different kinds of pipes, their materials and roughnesses epanetjs.com/blog/2026/07... 
Share your sessions with others.
Run /𝐬𝐡𝐚𝐫𝐞 and get an instant, easy-to-copy link.
- Best for reviews and debugging
- Anyone with the link can view the full session replay
- Your every step, every decision, every change
Done sharing? Type /𝐮𝐧𝐬𝐡𝐚𝐫𝐞 to disable the link instantly. 
Download Zed and view the full changelog here: zed.dev/releases/sta... Format-on-save is now disabled by default, except for languages with official formatters. For existing users, this means their custom configurations will stay, but any values that were previously "default" will be updated.
We've moved all the key agent-related settings to the settings editor. Configure LLM providers, external agents, and MCP servers directly from the Settings UI.
We now have a setting for rendering the current-line git blame in the status bar instead of inline at the cursor. Thanks Koshkaj!
🚀 Zed v1.10 is out!
Thanks to our friends at @hf.co we finally have support for llama.cpp in Zed. 
Full write-up, all 677 features and how it cleared the gold line
utm.io/urkYG None of this needed a smarter model. A 2018 XGBoost and a 2025 one score about the same.
What never scaled was the engineer running 20 feature loops a week. The agent runs 60 overnight for pocket change, and trying things was most of the job anyway. 
it didn't get there by brute force. Nobody told it crystals are periodic.
It worked out that raw atom-to-atom distances misread the chemistry across a cell boundary, then rebuilt them as metal-to-oxygen nearest neighbors.
677 features by the end. 😳 
The headline run was a materials-science task. Predict formation energy and bandgap of a transparent conductor from its crystal structure.
It landed 1st of 685. OOF score 0.04853 against a 0.05589 gold line. The humans who won it originally spent weeks. 
Sixty cycles later, across four MLE-bench datasets at once, it had two golds and three of four finishes above the Kaggle median.
No human in the loop the whole run. One NVIDIA L40 on @CoreWeave and a few dollars. 
Tabular ML didn't die. It was just waiting for agents.
We rented one GPU for the price of a coffee and asked an agent to top a Kaggle leaderboard on its own.
It came back with two gold medals. Here's the run 👇 
this is unexpectedly heartwarming github.com/Ranchero-Sof... Explore all 62 (and counting) here: hot-ar-summer.netlify.app/showcase 62 apps shipped at Hot AR Summer and counting. 🚀 Builders are turning out some seriously creative AR experiences. Browse the showcase and see what you could add to it! ✨
At PEGS Boston, Xiangxu Kong, PhD, Scientific Leader at Elicit, shared how protein sciences teams are applying AI to published knowledge to support engineering decisions.
www.youtube.com/watch?v=XEp... the options i see look like running bitwarden or hoping that proton doesn't enshittify
it has been fantastic, one of my favorite pieces of software. but they raised way too much money, the quality is going down, and they're doing dumb shit like putting logos on race cars
my bet: 1password will be unusable in 3 years
Setup is a toggle in workspace settings, and every policy change lands in your workspace audit logs.
Full breakdown of how each policy works in the docs ↓
docs.railway.com/enterprise/g... You can restrict deployments to approved sources: only GitHub orgs you allowlist can ship, and it's a hard boundary, even for admins Personal repos, CLI deploys, Docker images, and templates are all blocked, official database templates can still be deployed
