The scanner's download and execution hints do not establish a malicious chain: the registry self-update verifies tarball integrity, and no npm lifecycle hook is declared. The automatic refresh and use of a first-party Claude skill creates a guarded agent extension lifecycle risk, so warn rather than block.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review; previous stored version diff introduced dangerous source
Trigger
Starting the Claude host after the skill refresh interval expires, followed by a Claude session spawn.
Impact
Updated skill content can influence future Claude sessions; the inspected code does not establish a malicious payload or broader agent configuration change.
Mechanism
The host fetches skill files from raw.githubusercontent.com, writes them into its skill directory, and passes that directory to Claude with --plugin-dir.
Network endpoints1
raw.githubusercontent.com/lizard-build/skill/main