claude is eager, gpt-5 codex is surprisingly thorough hrm

i'll be honest the fact that this remains the only way to not get bloated up d.ts file is making me tempted to have my libraries stop exporting validation schemas directly https://pds.kelinci.net/xrpc/com.atproto.sync.getBlob?did=did:plc:ia76kvnndjutgedggx2ibrem&cid=bafkreibnyfywg3krciluonwss5nnrdmwp5hz5dnhrarg6gip2t2izjr54y

thought about [grfn](

GitHub

GitHub - TomerAberbach/grfn: 🦅 A tiny (~315B) utility that executes a dependency graph of async functions as concurrently as possible.

🦅 A tiny (~315B) utility that executes a dependency graph of async functions as concurrently as possible. - TomerAberbach/grfn

) and it turns out all i wanted was just a function that unwraps promises and passes it over to a callback

can't tell if ublock origin on chrome is just borked or if this filter rule never worked on chrome version to begin with but the $replace filter syntax doesn't work

im sure there's a bigger car export i could take a look at but maybe this is enough https://pds.kelinci.net/xrpc/com.atproto.sync.getBlob?did=did:plc:ia76kvnndjutgedggx2ibrem&cid=bafkreibz34xlwewog7iobbwiii7mhdy6fjlo77lbhfddp3xi7dudlnsmiy

hrm, may or may not be a mistake RE:

Bluesky Social

mary🐇 (@mary.my.id)

....maybe i could just have a script that grabs a tarball of the repo instead, hrrmmmmm

https://pds.kelinci.net/xrpc/com.atproto.sync.getBlob?did=did:plc:ia76kvnndjutgedggx2ibrem&cid=bafkreicoo4sbtsvfrnshr5lmyknn6prn626x6ikj5aauskxf636s7icmru

time has come when i finally have to use git submodules

whoa hello another xaev song [youtu.be/K_tbRhPaSkI]( ) [MRX-168 Beyondcore ESSENCE 268...]( )

a guide for migrating an account from a dead/borked PDS to a new one would be useful i think hrmm

what is dry and wet claude supposed to be all i can think of is golden gate claude