registry / Credential exfiltration / @ricsam/r5d-browser / 0.0.189-prod.1790576235476.ga922cc6c91ba099b9eb51a1018d554e54207deb3
@ricsam/r5d-browser@0.0.189-prod.1790576235476.ga922cc6c91ba099b9eb51a1018d554e54207deb3
`r5d-browser start --auto-update` enables automatic client updates for this environment and state directory. The setting persists; `--no-auto-update` disables it. The client checks the authenticated environment's selected release once a minute and stages
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThe explicit CLI activation and package-aligned default reduce stealth concerns, but HTTPS alone does not restrict a reusable credential to its intended service. The unrestricted origin selection plus token forwarding warrants a non-blocking warning.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
The user runs r5d-browser start with a selected or configured origin.
Impact
A configured non-r5d endpoint could receive the service credential and control the connected browser session.
Mechanism
The client forwards a bearer token to the selected relay and dispatches returned browser actions, including supplied page JavaScript.
AI rationale
The explicit CLI activation and package-aligned default reduce stealth concerns, but HTTPS alone does not restrict a reusable credential to its intended service. The unrestricted origin selection plus token forwarding warrants a non-blocking warning.
Evidence
dist/mjs/personal-main.mjsdist/mjs/browser-runtime.mjs
Network endpoints1
r5d.dev
Decision evidence
public snapshotAI called this Suspicious at 91.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning