AI Security Review
scanned 2h ago · by lpm-firewall-aiLPM blocks this version under the AI-agent control-surface policy. npm postinstall installs a package-supplied Claude Code skill into the user's home directory and alters its configuration. It then executes a helper that downloads catalog content into that agent-controlled skill tree.
Decision evidence
public snapshot- package.json defines postinstall: node install.js.
- install.js runs automatically without a setup argument.
- install.js copies package files into ~/.claude/skills/hyper-animator.
- install.js creates ~/.claude/skills/hyper-animator/.env.
- postinstall executes sync-catalog.py through child_process.execSync.
- sync-catalog.py invokes hyperframes add and stores fetched HTML in the Claude skill tree.
- No credential value is read or exfiltrated by install.js.
- MiniMax API requests are in an explicit generation helper, not install.js.
- No eval, VM loading, or shell interpolation beyond the fixed Python command was found.
- The copied content is package-branded, but it still mutates a foreign Claude control surface automatically.
Source & flagged code
4 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 unpkgInstall-time source drops package-supplied AI-agent/MCP control files or instructions.
install.jsView on unpkg · L12Package ships non-JavaScript build or shell helper files.
skills/hyper-animator/scripts/lint-html.pyView on unpkg