AI Security Review
scanned 37m ago · by lpm-firewall-aiNo confirmed malicious attack surface was established. The package is a deployment CLI with user-invoked scaffold, resolve, apply, adopt, demo, and tunnel commands that perform expected filesystem, SSH, Docker, Cloudflare, Forgejo, and webhook operations.
Decision evidence
public snapshot- dist/init/init.js runs git/pnpm and writes scaffold files, but only from explicit init command
- dist/resolve/resolve.js dynamically imports user config path and reads Cloudflare token env for zone discovery
- dist/demo.js can run docker/git-like commands, write .demo/desired-state files, and call Cloudflare/Forgejo APIs, but only via demo up/clear
- dist/apply/apply.command.js posts to a user-supplied Discord webhook from resolved graph outputs
- package.json has no preinstall/install/postinstall lifecycle hooks
- dist/cli.js only dispatches explicit CLI commands via @stricli/core
- No import-time harvesting or outbound request observed in package entrypoints
- Network endpoints are deployment/product-aligned: Cloudflare API, configured Forgejo/domain, local demo services, optional Discord webhook
- No AI-agent control-surface writes or persistence outside user-requested deployment/scaffold flows found
- File writes are scoped to user-selected project dirs, desired-state artifacts, known-hosts/secrets files, demo state, or remote deployment paths
Source & flagged code
4 flagged · loading sourcePackage source references child process execution.
dist/init/scaffold-app.jsView on unpkg · L1Package source references dynamic require/import behavior.
dist/resolve/resolve.jsView on unpkg · L6A single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/demo.jsView on unpkg · L29Source combines command execution, command-output handling, and outbound requests; review data flow before blocking.
dist/demo.jsView on unpkg · L29