The install hook registers the package's own agent skill by default for direct dependencies, which meets the warning-level agent extension lifecycle risk. The skill content inspected is package guidance, and no malicious agent steering or data collection was identified.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
Package installation as a direct dependency, unless skipped by the install conditions or disabled in project configuration.
Impact
This adds or refreshes an agent skill without a separate setup command, creating an agent extension lifecycle risk.
Mechanism
The postinstall reads the project setting and writes the bundled Ax skill and guide files to .claude/skills.
Evidence
package.jsonscripts/postinstall.mjsscripts/skill-config.mjsscripts/skill-install.mjsskills/ax-llm.md.claude/skills/<skill.name>/SKILL.md.claude/skills/<skill.name>/references/<guide>