@swyx I think there’s maybe just an angle of AI native homelab here. Good homelab with correct secure layered setup is not a new setup. But having agents/sem LLM gateway inside, and skills to operate it from the outside is the elevator
@helloiamleonie That’s how we ended up building a platform that is deliberate about being agnostic to these. Instead, we define interfaces harnesses need to support for runtime, evals and skills and let them be. The platform takes care of security scoping, execution env, triggers and outputs
@AlexPalcuie To be fair, your current office is our previous office and it’s one of the worst designs on room-to-desk ratio. Hope the new one plays out better for you
@thekitze Exactly the same setup here 👋 I have a separate travel kit always ready to go, also with Chromecast. It’s been such a change when all the devices just auto connect and internal homelab services are all just there
Great morning bringing the speakers from @aiDotEngineer to Downing Street to discuss transforming the state.
Through the Incubator for AI and the No10 Innovation Fellowship, we are making sure that top AI talent can help build a better Britain!
https://t.co/JWMA3OddnT
What a day! As part of the @aiDotEngineer conference, I was invited to an event at No 10. There are plenty of practical problems to solve, and it’s encouraging to see the Government actively seeking advice from industry leaders.
Comprehensive context creates maintenance burden. Staleness creeps in, accuracy drops, no one owns it. Curated context can be kept current, validated, owned. IDP is your friend here.
Teams ask "how do I give Claude context?" Instinct: dump repos, link all Confluence, create massive https://t.co/BcmRSxfpGJ. Wrong. More comprehensive ≠ better. Quality over quantity.
MTTR improved once context existed. Debugging sped up when Claude could query actual architectural decisions instead of guessing from training data. Context infrastructure pays off in metrics.
The onboarding agent pattern: configure Claude NOT to generate code, but to guide learning. Pull context from Confluence + GitHub + Jira, generate bespoke plan. That's the cycle working.
Infrastructure work isn't flashy. Building knowledge bases, auto-generating docs with quality gates, creating skills. It's not free. That's what makes coding agents work in enterprise.