Skip to main content
Point Alice.In at a repository and it infers the shape of your system: services, libraries, local tools and the external services you depend on. The result is a map you can walk, not a diagram someone drew once and forgot.

First run

The map builds itself in front of you the first time — watch it construct. It takes about a minute. You do not draw anything. Alice.In reads the code and infers the structure: what is a service, what is a library, what is a tool you run locally, and what you depend on from outside.

Reading it

  • Modules are laid out in districts, with the external services you call sitting on the outer ring.
  • Edges are the connections between them, typed: HTTP, MCP, OTEL, WS, DEP.
  • Tiers distinguish a service from a library, a local tool, or something external.

Peeling it back

Behind the gear you can toggle each connection type and each module tier on and off. A map showing everything at once shows nothing, so turn layers off until only the question you are asking is left. Your choices persist between sessions. Hover or click a module and only its edges label themselves — no global clutter.

The module panel

Click any module for what it is, what it connects to, which files belong to it, and what is happening inside it now. Ask a question there and the answer comes back grounded in the real files, with receipts.

Flows and the walkthrough

  • Flows are curated stories of how data actually moves through your system, played out on the map.
  • The walkthrough is a guided first tour. Dismiss it any time; it will not come back uninvited.

Recent work shows on the map

Once you have connected your agent, the commits it records are matched against your modules by the files they touched. A module that has been worked on recently reads as active, and one worked on in the last day reads as fresh. Over a week, the map shows you where the effort actually went — which is not always where you thought it went.