Source inspection confirms an install-time setup path that writes package-owned agent and skill assets into AI harness directories, with GitHub Copilot selected by default when no interactive terminal is available. This supports a warning for agent extension lifecycle risk, not a confirmed malicious attack.
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
npm installation runs the postinstall hook when lifecycle scripts are enabled.
Impact
Package-provided agent and skill instructions become available to the selected harness; no additional attack behavior was established from this setup.
Mechanism
The hook calls harness setup, which defaults to GitHub Copilot without an interactive terminal and writes the bundled assets into harness directories.
Evidence
package.jsonscripts/postinstall.mjssrc/harness-setup.mjssrc/install-user-assets.mjssrc/harnesses.mjs