is there any open source skill benchmark workbench for claude code? i want to define multiple versions of a skill, acceptance criteria (to be evaluated by a separate agent), and a runner that does repeated runs of different versions of my skill to see if there’s a statsig improvement in any version
seems like on ios, there’s a regression: if you go from feed to profile, back swipe no longer works reliably if you start it from a post. also quote embeds seem to maybe interfere with back swipe as well
i don't remember why i made this https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:fpruhuo22xkm5o7ttr2ktxdo&cid=bafkreigu6iusqnlay75fnsp3gmhlkhg5kmwrqtyzjt7xb4i47ma64kn2vu
wooooo! RE:
hmm i started seeing feed context hints in stories, is that expected for my account or did it accidentally get turned on for everyone RE: View quoted note →
i’m team em dashes for life the actual tell of most AI content is just that it doesn’t have anything to say and is boring to read people who can’t tell good writing from bad writing by content shouldn’t have any sway over how we write RE:
digging through old react commits, my favorite titles so far
ok this is actually pretty cool RE: https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:fpruhuo22xkm5o7ttr2ktxdo&cid=bafkreidxyfco5tyjkoxvthpl3rwuqrcc3yerwibt67w4aoiryq7liudpqm View quoted note →
ah well. okay then https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:fpruhuo22xkm5o7ttr2ktxdo&cid=bafkreifxtx2rptuc3avgjgahu64imq2thhnud2ovloqd522gn7u6nzxefm
working on a skill to teach claude to export archeology of a git repo into deciduous