@twing/cli@1.3.23
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe lifecycle behavior is a guarded, first-party agent integration setup, not evidence of credential theft or remote code execution. Because it automatically changes global agent configuration and enables later hook execution, residual lifecycle risk warrants a warning.
Decision evidence
public snapshot- The postinstall hook runs machine setup for global installs or when explicitly passed the machine-setup flag.
- Machine setup wires Claude, OpenCode, and Codex integrations and marks the installation as managed.
- The setup writes a persistent Claude resolver, an OpenCode plugin loader, and Codex hook configuration; these integrations can run in later agent sessions.
- Ordinary project installs exit before machine setup; Node version is also checked before loading setup.
- The integrations are first-party Twing components, and the source describes the global install as deliberate machine setup.
Source & flagged code
13 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 unpkgPackage source references child process execution.
dist/opencode-adapter.jsView on unpkg · L12Findings
Affected versions and remediation
This report applies to @twing/cli@1.3.23.
See all versions of @twing/cli for other recorded verdicts.
- Review the evidence and your use of @twing/cli@1.3.23 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.