Source inspection found first-party agent integration setup activated by explicit plugin commands, including a Codex instructions-file write. The postinstall hook only prints guidance, so this supports a warning for agent extension setup rather than a publish block.
Static reason
One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review
Trigger
A user runs the plugin add or update command.
Impact
The installed integration can influence later coding-agent behavior; no unconsented install-time mutation or separate attack behavior was identified.
Mechanism
The command invokes client-specific setup, including writing the package skill into Codex instructions and registering or updating the Claude Code plugin through its CLI.
Evidence
dist/commands/codex.jsdist/commands/claude-code.jsdist/commands/plugin.jsCodex instructions file