The reachable postinstall performs first-party agent skill setup in the consumer project, which warrants a warning under the agent integration policy. Binary entrypoint files remain unresolved, but the inspected setup source establishes a concrete warning without evidence for a block.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
Package installation outside CI when automatic setup is not disabled in project configuration or through environment flags.
Impact
The project’s Claude skill set changes during dependency installation; the inspected source does not establish broader agent-setting changes or data transfer.
Mechanism
The postinstall hook calls the bundled installer, which writes the Ax skill and its guide files under .claude/skills.
Evidence
package.jsonscripts/postinstall.mjsscripts/skill-config.mjsscripts/skill-install.mjs.claude/skills