@viraatdas/rudder@3.5.4
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 4d ago · by lpm-firewall-aiThe postinstall hook performs Rudder’s workspace-tool setup, while reviewed credential flows use Rudder Cloud for account comparison and ChatGPT for Codex usage lookup. The flagged agent paths correspond to cloud workspace setup or local data reads; no concrete attack was identified.
Decision evidence
public snapshot- The postinstall hook skips when disabled or running as root, then invokes Rudder’s Jujutsu setup; its CLI entrypoint is Rudder’s own command.
- Cloud credential checks read local account identity data and compare it with credentials returned by Rudder Cloud; the reviewed callers are cloud commands and workspace setup.
- Quota lookup reads Codex’s saved access token and sends it to ChatGPT’s usage endpoint; the report is fetched on demand or by the board quota handler.
- The reviewed agent-related paths stage Claude credentials into a cloud workspace, locate local Claude conversation records, and read Codex’s local model cache.
- The other inspected runtime excerpts cover local board handling, text extraction, redaction, improvement collection and scheduling, backends, run startup, version checks, feedback, and Jujutsu setup.
Source & flagged code
21 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time source downloads a native archive from a fixed external host without transport verification, extracts it, and installs an executable payload.
package.jsonView on unpkg · L1Findings
Affected versions and remediation
This report applies to @viraatdas/rudder@3.5.4.
See all versions of @viraatdas/rudder for other recorded verdicts.
- Review the evidence and your use of @viraatdas/rudder@3.5.4 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.