The postinstall hook automatically overwrites an already installed agent skill, which creates an install-time agent extension lifecycle risk. The generated file is limited to this CLI’s own skill, so the evidence supports a warning rather than a block.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review; source fingerprint signature matched known malicious package; routed for review
Impact
Agent instructions can change during package installation without a separate user command; the changes are limited to this package’s own skill.
Mechanism
The hook checks supported agent skill locations and overwrites an existing automations-cli/SKILL.md with generated content.
Evidence
package.jsonscripts/refresh-skills.mjsdist/chunk-LNPR6Q65.js~/.claude/skills/automations-cli/SKILL.md~/.codex/skills/automations-cli/SKILL.md~/.copilot/skills/automations-cli/SKILL.md~/.cursor/skills/automations-cli/SKILL.md~/.gemini/config/skills/automations-cli/SKILL.md