The postinstall hook automatically installs the package’s own agent skills and registers its own MCP server, which meets the warning criteria for agent extension lifecycle risk. I found no evidence of permission hijacking or unrelated credential exfiltration, so this does not warrant a publish 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; previous stored version diff introduced dangerous source
Trigger
npm install, through the package’s postinstall lifecycle script.
Impact
Agent sessions discover the package’s skills and MCP server without a separate setup command; this is an agent extension lifecycle risk.
Mechanism
The hook runs the CLI to install bundled skills and register dr-zendesk-kb with Claude Code.
Evidence
package.jsonscripts/postinstall.mjsdist/cli.js~/.claude/skills/<name>/SKILL.md~/.gemini/skills/<name>/SKILL.md~/.gemini/GEMINI.md~/.claude.json