@chemx/create-chemx@26.9.17-606
Chemical X Protocol: Private drop-in architecture starter kit and capsule generator
AI Security Review
scanned 11d ago · by lpm-firewall-aiThe automatic postinstall writes MCP configuration into consumer AI editor control surfaces without consent. This is concrete install-time agent control-surface mutation and warrants blocking.
Decision evidence
public snapshot- The package runs cli/postinstall.js automatically during installation.
- That hook calls installAllMcpConfigs for the consumer project without a consent check.
- The installer creates and writes a Cursor MCP configuration during postinstall.
- The installer also configures VS Code and Antigravity MCP integrations.
- A separate installer function appends agent directives to an existing AGENTS.md when invoked.
- The MCP installer test confirms the Cursor configuration is created.
- The postinstall skips installation when run inside the package itself.
- The MCP configuration registers a named Chemical X server; the inspected install path shows no credential collection or unrelated network transmission.
Source & flagged code
10 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgPackage source references child process execution.
cli/audit-preflight-git.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @chemx/create-chemx@26.9.17-606.
See all versions of @chemx/create-chemx for other recorded verdicts.
- Avoid installing @chemx/create-chemx@26.9.17-606. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.