@viraatdas/rudder@2.15.1
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 9d ago · by lpm-firewall-aiSource inspection found no concrete malicious behavior. The install hook performs a bounded jj setup, and the credential-bearing cloud transfer is part of an explicit Rudder Cloud command.
Decision evidence
public snapshot- package.json runs scripts/postinstall.mjs; the hook checks for jj and installs it through brew or cargo only if it is missing.
- Cloud launch is reached through cloud subcommands, stages a snapshot, and posts it to the Rudder Cloud control plane. Claude credentials are staged for that cloud workflow.
- The handoff and model locations inspect local Claude project conversations and read the local Codex model cache; they do not mutate agent configuration.
- The improvement scheduler writes a launch agent only for its install action; collected task summaries are redacted and trimmed.
- The reviewed runtime sources show a loopback-bound local board, a package-version lookup, optional feedback reporting, and a separate worker process.
Source & flagged code
18 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@2.15.1.
See all versions of @viraatdas/rudder for other recorded verdicts.
- Review the evidence and your use of @viraatdas/rudder@2.15.1 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.