← Downloads
Local Shadow · Prerelease 1.5.0-shadow.8

Configure your ROS 2 interfaces

Select your actions or velocity topics, identify the receiving node, and prepare a local Shadow comparison.

Files are processed in this tab. Their contents are not uploaded or saved by this page. No robot commands are sent.

  1. 1Import
  2. 2Interfaces
  3. 3Configuration
  4. 4Export
Interface configuration

1. Import

What will I need, and what will I get?

Start with one command boundary in an isolated simulation. Prepare the interface catalog, one real message/goal example, the robot description, and the controller/configuration or calibration files relevant to that command. For a velocity topic, identify which node receives it.

No repository login, remote-machine access or cloud account is required. You review a separate workspace before running it. Setup time depends on whether your message semantics and configuration sources already fit; there is no measured customer setup-time estimate yet.

The installed CLI evaluates locally without a cloud connection and writes a readable report, per-check results and Evidence. This web page needs internet to load; the CLI, downloaded settings and installed documentation can be used offline after dependencies are installed. ROS graph discovery still uses local DDS traffic.

Choose one change together, such as a controller configuration or calibration copy. Keep the approved baseline unchanged, collect both observations, and compare them. Never change a live robot just to trigger a block.

WOULD_ALLOW means the declared authorization checks matched; it does not mean the motion is safe. Existing motor limits, readiness, emergency stops and collision checks remain in place. Public case studies, research plans and contribution credit need a separate agreement.

On the machine with ROS 2 and your interface packages sourced, use an isolated or simulated graph.

Discover your interfaces

Install the separate evaluation package, then collect action/topic metadata and installed definitions. No command is published.

Get evaluation 1.5.0-shadow.8
rlsok profile discover --output catalog.json

Reuse a saved setup

Restore interface choices, mappings, example goals and expected configuration. Reselect private fact files before exporting.

Importing a new catalog starts a new setup. A saved catalog is a snapshot; capture fresh observations before evaluation.