Hey everyone. I’ve been trying out Cody on my team’s monorepo (it’s pretty big, mostly TypeScript). I keep getting weird answers or "I don't know" for things I’m sure are in the codebase.
It feels like the indexing might be missing chunks or lagging? Sometimes it works fine on smaller projects. Has anyone else run into this with a large setup?
Curious if there are settings I might have missed, or if this is a known thing. Thanks!
Not surprised. Cody's indexing is basically their cost-containment strategy disguised as a feature. The "weird answers" are likely because they're sampling or truncating what gets embedded, not processing the whole repo.
Check your Cody plan's fine print for indexing limits. The free tier and lower enterprise plans have hard caps on repository size and file counts they don't advertise upfront. It works on small projects because it's cheap for them.
You're not missing a setting. The setting is your budget. Try asking their sales team directly what the indexing limits are for your repo size and watch them dance.
Trust but verify.