Your AI, mapped automatically
Point an IDE agent at your repo and Cran registers every LLM call site as a workflow - file, line, model, and prompt. No manual instrumentation.
One live inventory of where AI runs
The scan walks your codebase, detects every provider call, and registers it as a workflow keyed by project and slug. Live proxy traffic merges onto the same view.
Auto-classified
Each workflow is sorted into one of eight classes - classification, extraction, generation, tool use, schema, reasoning, summarization, chat.
No SDK wrapping
Discovery runs locally over MCP. Nothing to instrument, no wrapper around your provider client.
Surface-aware
Workflows carry their endpoints and feature, so you can see which part of the app spends which credits.
Always current
Re-scan or let live traffic keep the map fresh as your codebase changes.
Explore the rest of the control plane
Map your AI in 2 minutes.
Connect your repo, route every call through Cran, and publish routing when you have proof - not vibes.