Most of our usage is now coming from agents so to monitor how they use and mention our tools, with @DAKlingbeil
, we're now running regular 10k queries to all the most popular coding agents, share the traces on @huggingface and analyze the results (ex https://
huggingface.co/datasets/clem/
hf-coding-tools-traces
…).
CODE
-
Monitoring AI Agents Tool Usage Through 10k Query Traces
By
–
-
LangSmith Observability Reveals Agent Development Patterns Across Billions Runs
By
–
We analyzed this information from LangSmith Observability data across billions of agent runs, and we’re just getting started. Stay tuned for more LangSmith Signals as we share how devs are building agents, by the numbers.
-
API Security Breach Exposes Public Project Chat Data
By
–
Crazy. “We also retroactively patched our API so public project chats couldn't be accessed, no matter what. Unfortunately, in February, while unifying permissions in our backend, we accidentally re-enabled access to chats on public projects.”
-
Claude Cowork: Free AI Agent Tool for Business Professionals
By
–
NEW: I just dropped my free Claude Cowork in 5 Minutes quick start for business professionals who want to use AI agents without the terminal. Cowork is made for less technical, everyday business work. It lives inside the Claude Desktop app and works by pointing it at a
-
Chronicle: Codex gains experimental memory and context awareness
By
–
Chronicle is an experimental feature giving Codex the ability to see and have recent memory over what you see, automatically giving it full context on what you're doing. Feels surprisingly magical to use. https://t.co/JnDl5Bk17o
— Greg Brockman (@gdb) 20 avril 2026Chronicle is an experimental feature giving Codex the ability to see and have recent memory over what you see, automatically giving it full context on what you're doing. Feels surprisingly magical to use.
-
Google DeepMind Forms Strike Team to Compete with Anthropic
By
–
Google has formed a dedicated strike team within Google DeepMind to improve its AI coding models, driven partly by pressure from Anthropic, whose tools are seen internally as more advanced. Leadership, including Sergey Brin (!), is pushing urgently toward “agentic” AI systems
-
Deploying Long Horizon Agents: Infrastructure Guide
By
–
how to deploy long horizon agents, and all the infra you need!
-
Replit Agent Now Supports Mobile Apps and Slide Decks in Existing Projects
By
–
You can now add a mobile app, slide deck, or animation to any web app you have built prior to Replit Agent 4.
— Replit ⠕ (@Replit) 20 avril 2026
Everything lives in the same project, with shared branding & data.
Just open your app and ask the Agent what you want to build. You also get $5 in credits to get… https://t.co/n0rx4H7SjiYou can now add a mobile app, slide deck, or animation to any web app you have built prior to Replit Agent 4. Everything lives in the same project, with shared branding & data. Just open your app and ask the Agent what you want to build. You also get $5 in credits to get
-
Six Dimensions Aligning Data Science Success Framework
By
–
Data Science works when 6 dimensions align: Goals (value, decisions)
Methods (stats, ML/DL, A/B, viz)
People (DS+ML+Biz+Domain)
Processes (collect→clean→train→deploy→monitor)
Tech (Python/R, TF/PyTorch, cloud, SQL/NoSQL, BI)
Culture (collab, ethics, learning, -
Kimi K2.6 Tops SWE-Bench Pro, Beating Claude Opus 4.6
By
–
Kimi K2.6 just dropped. And it crushed Claude Opus 4.6 on SWE-Bench Pro. – Kimi K2.6: 58.6
– GPT-5.4 xhigh: 57.7
– Gemini 3.1 Pro: 54.2
– Claude Opus 4.6: 53.4 An open-source model is now #1 on agentic coding. The moat around frontier labs is shrinking fast.