Introducing Anlyon Integrations.
Connect Stripe, GitHub, Notion, Intercom and more with one key.
Your agent gets every read, and writes that run under your limits, wait for you when they should, and come back with a receipt.
It never holds the key.
shipped hosted functions on anlyon today
write a multi step task in JS or Python (read the order → check it → refund → email).
It runs with no network and no keys and every write waits for a yes before it goes out.
https://t.co/iF7bOzaqOq
I heard some version of this question on almost every call dueing the beta.
No body asked if the agent was smart enough they asked what happens when it's wrong in customer's account.
Slop machines know no rest.
v0.4.5 ships major improvements to lens corrections.
RAWmakase now applies lens corrections embedded in RAW files by cameras: distortion, chromatic aberration, and (for many models) vignetting.
No need to import Adobe *.lcp profiles anymore - but you still can, if you want.
Full release notes: https://t.co/y037KVBlCz
Yesterday's release (v0.4.4):
https://t.co/fbtU6qpBL8
If you notice any bugs, open a detailed issue on GitHub. Don't complain here 😉
Next, back to work on:
- built-in film simulations and presets
- built-in camera profiles
- missing file management tasks
The highlights/shadows parity research agents have been going in circles, making minimal progress. Need to try a different approach. Or wait for Opus 6.5.
Building something people actually use or would like to use is hard.
You really have to let go your ego of being right and build what people actually want
Railway: "rotate all your secrets."
Also Railway: "deploys are paused."
Your keys were leaked and you couldn't fix it.
Keep provider keys out of your env. @anlyonHQ runs the API call and the app only holds a token you can kill instantly.
As if PHP2Shell wasn't enough for one day... now GitLab. 😅
🚨 CRITICAL GITLAB ALERT (CVSS 10.0)
CVE-2026-110264 could allow unauthenticated RCE on self-managed GitLab servers.
⚠️ Potentially affects 12 years of releases, dating back to 2014.
Patches available. Update immediately!
https://t.co/MtIwEFk7QY
#gitlab #cybersecurity
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent.
It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
🚨 LiteLLM, a Python package for connecting to AI model providers with 92M+ monthly downloads, has been compromised.
Version 1.107.0 ships credential-stealing malware through PyPI wheels and Docker images, targeting cloud credentials, developer secrets, and AI tools.
most "human in the loop" is a rubber stamp you approve a summary click yes 40 times a day and the request can change after you approved it a yes should be bound to the exact call
every day there is a new AI agent setup that promises to replace an entire engineering team
meanwhile I am still trying to get one agent to stop overengineering a simple feature