@viraatdas/rudder@2.16.6
A TUI for managing coding agents: run Claude Code, Codex and opencode side by side, each in its own isolated jj workspace.
AI Security Review
scanned 7d ago · by lpm-firewall-aiThe install hook only checks for and conditionally installs jj. The reviewed cloud, telemetry, updater, and scheduling behavior is tied to Rudder features; no concrete malicious behavior was established.
Decision evidence
public snapshot- The install hook checks for jj and only invokes Homebrew or Cargo to install it when missing.
- Cloud snapshots, including Claude credentials and selected environment values, are built in an authenticated Rudder cloud launch path and sent to its launch API.
- The agent-related helpers use local Claude conversation storage and read Codex's local model cache.
- Improvement collection reads registered project runs; scheduling is an explicit action, and the default autonomy is propose.
- The board checks loopback hosts before serving local controls.
- Update checks have disable flags and skip CI or development installs; telemetry is disclosed at command startup and can be disabled, and feedback text is redacted.
Source & flagged code
17 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
dist/util.jsView on unpkgFindings
Affected versions and remediation
This report applies to @viraatdas/rudder@2.16.6.
See all versions of @viraatdas/rudder for other recorded verdicts.
- Review the evidence and your use of @viraatdas/rudder@2.16.6 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.