@viraatdas/rudder@2.14.31
⚠ Under reviewA Claude Code-style terminal app for running coding agents with worktree-isolated runs.
AI Security Review
scanned 15d ago · by lpm-firewall-aiInspected behavior did not establish a concrete attack: postinstall provisions jj, and the cloud code stages credentials as part of workspace operations. The required clean citation coverage exceeds the eight-citation limit, so the result remains unresolved under the supplied review rules.
Decision evidence
public snapshot- The package has a postinstall hook that probes for jj and attempts installation through Homebrew or Cargo.
- Cloud snapshot creation can stage credentials when includeCredentials is enabled; the inspected caller sends the snapshot to the Rudder workspace attach endpoint.
- The inspected agent migration path prompts for confirmation in a TTY before proceeding when agents would be migrated.
- The required clean citation set contains 14 paths, exceeding the limit of 8 citations; no concrete attack behavior was established in the inspected source.
- The inspected handoff code locates Claude project conversations, and the inspected models code reads a local Codex model cache.
- The postinstall behavior inspected installs jj and exits; no malicious payload or foreign agent configuration mutation was established.
Source & flagged code
21 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@2.14.31.
See all versions of @viraatdas/rudder for other recorded verdicts.
- Review the evidence and your use of @viraatdas/rudder@2.14.31 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.