Inspected source shows that postinstall only prints a reminder, and agent integration setup is explicitly invoked and package-owned. The flagged download and native-loading code is consistent with model setup and platform-specific database bindings, with no concrete attack behavior identified.
Static reason
One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review; previous stored version diff introduced dangerous source
Trigger
Package installation prints a reminder; agent configuration changes require an explicit plugin setup command.
Impact
No confirmed attack surface was established from the inspected source.
Mechanism
Postinstall outputs a message. Explicit setup commands register the package’s own integration, while model files are downloaded from Hugging Face into model storage.
Evidence
scripts/postinstall.jsdist/commands/claude-code.jsdist/commands/codex.jsdist/commands/plugin.jsdist/commands/setup.jsdist/lib/workers/download-worker.js