Comment on Interview with Boris Cherny [video]Comments−albert_e1moSo maybe SKILL.md and CLAUDE.md should have model version pinned?Somewhat like AWS CloudFormation templates start with something like TemplateVersion: 2017-10-01But maybe that's old school thinking still.−crazylogger1moThey should just be symlinks to CONTRIBUTING.md and other docs humans are meant to read.The whole idea behind LLM is they use the same language and tools as us, so we literally don't design anything specifically for them.−knighthackerOP1moI'm thinking a lot about this recently as I'm building AQ.dev (Multiplayer coding harness). My instinct is to keep it simple and not trying to do too many shenanigans with these files and let engineers maintain control.−verdverm1mothis assumes everyone uses the same model, but where I'm at, we use a wide variety of modelswe keep the AGENTS.md simple and monorepo quirk focussed, no instructions for specific agents or models
Comments
So maybe SKILL.md and CLAUDE.md should have model version pinned?
Somewhat like AWS CloudFormation templates start with something like TemplateVersion: 2017-10-01
But maybe that's old school thinking still.
They should just be symlinks to CONTRIBUTING.md and other docs humans are meant to read.
The whole idea behind LLM is they use the same language and tools as us, so we literally don't design anything specifically for them.
I'm thinking a lot about this recently as I'm building AQ.dev (Multiplayer coding harness). My instinct is to keep it simple and not trying to do too many shenanigans with these files and let engineers maintain control.
this assumes everyone uses the same model, but where I'm at, we use a wide variety of models
we keep the AGENTS.md simple and monorepo quirk focussed, no instructions for specific agents or models