Loved working with the @github Security Lab and fellow maintainers of widely used open-source projects! Amazing collaboration, and security insights into open-source .
I highly recommend this program to maintainers looking to improve their projectsโ security. GSOF Session 4!
Proud to have been part of GitHub Secure Open Source Fund, Session 4 with @caracalrun.
Learned a lot from the @github Security Lab, fellow maintainers, and the incredible people behind the program. Grateful for the opportunity to build more securely.
https://t.co/5f0BC5kkws
Caracal is now live on Product Hunt!
https://t.co/tT30dP2Edl
Caracal gives AI agents authority, not credentials: short-lived, scoped, instantly revocable access to tools, APIs, and MCP servers. Every action is approved by policy before it runs.
Been using Fable 5 and GPT 5.6 Sol for the past few days. Accuracy is pretty much the same, but Fable is 5โ6ร faster. For my workflow, speed matters more, so I'm team Fable. ๐
Every AI company seems to be solving agent authorization differently.
If your AI agent needs access to a production database or internal/third party API...
What's your current approach?
I'm genuinely curious how people are handling this today.