Explain a large repository
Drop into an unfamiliar codebase and ask. Sarvam Code maps the structure, the key modules, and how a request flows through the system.
› I just cloned this repo. What is it, and where does an order get charged?
• Let me map the project before answering.
• Ran rg --files -g '*.go' | head
└ cmd/server/main.go
internal/checkout/handler.go
internal/payments/stripe.go
internal/orders/repo.go
• It's a Go checkout service: an HTTP server (cmd/server) over three domains: checkout, payments, orders.
• An order is charged in internal/payments/stripe.go:Charge, called from checkout/handler.go once the cart validates. Want me to walk that path?
- 01Open the repo and ask what it does or where a feature lives.
- 02The agent builds a map of modules, dependencies, and conventions.
- 03It answers with the specific files and call paths that matter.
- 04Follow up in the same session as you go deeper.
Onboarding onto a new codebase is mostly reading. Sarvam Code reads it the way you would, then gives you the short version: what the major pieces are, how they fit together, and where to start for the change you have in mind.
Repo Q&A is exactly the kind of read-heavy work where an affordable model shines: the agent reads far more than it writes, and those input tokens cost a fraction of a cent. You get oriented fast, without spending frontier prices to do it.
Curious what it costs to run work like this? See cost per task, or the full pricing.