Skip to main content
Shared architecture maps for software teams

Share how your system works.
See what changed.

Silhouette Canvas turns a local repository scan into an interactive architecture map. Share one link with your team or external collaborators, then compare revisions to see how the system changed.

View an example map

Paste it into a repository-aware AI coding agent. The prompt scans locally and publishes only bounded architecture metadata, never source code.

One map for the full system

RESTWebSocketWorkersStateUI
Shared architecture

Every part of your system, mapped in one place

Repositories, APIs, workers, state, integrations, and UI meet in one readable architecture map. Canvas gives every teammate the same system context and preserves how it changes.

19 components · 23 flows

How it works

Scan locally. Share the map. Compare every revision.

A repository-aware AI creates and validates the architecture summary locally. The prompt publishes that bounded artifact, Canvas renders it, and later scans become a clear visual before-and-after.

AI coding agent / local repository

$ You are an AI coding agent working inside this repository.

  1. Before inspecting the repository or creating files, ask the user to choose the privacy mode. Your first response must be exactly:
  2. "Canvas can analyze this repository locally without sending its source code anywhere. Publishing later sends a sanitized architecture summary to https://canvas.silhouette.exchange/api/scans; the resulting map is public, unlisted, and expires after 90 days. The upload contains component names, kinds, relationships, and declared runtime metadata, but no source code, file contents, source paths, endpoints, evidence paths, owners, secrets, environment values, customer data, or absolute paths. Reply `publish` to authorize that one HTTPS upload after local validation, or `local` to keep every generated file on this machine."
  3. Wait for the user's explicit `publish` or `local` reply. Do not interpret the pasted prompt itself as upload authorization. After the user chooses, do not ask the same conversational consent question again. A native tool or sandbox approval for the HTTPS request may still appear and must be handled through its normal permission mechanism.
  4. Analyze the repository and create two local files for Silhouette Canvas:
  5. - .silhouette/scan.json: the validated architecture data
  6. - .silhouette/scan.md: a portable Markdown summary with a Mermaid flowchart
  7. ready to paste into your AI coding agent
copy promptpaste into your AIreceive the map link
01

Create the local scan

Choose local or publish mode, then let your repository-aware coding agent build the rich scan on your machine.

02

Publish the shared map

Publish mode removes local paths, endpoints, owners, and evidence before returning one public, unlisted link.

03

Compare the next revision

Update the same map from a later scan to see added, removed, and changed components and flows.