AI Security Review
scanned 15h ago · by lpm-firewall-aiReview flagged AI-agent configuration or capability changes. This remains warn-only unless evidence shows foreign-agent hijack through preinstall/install/postinstall, hidden persistence, exfiltration, remote code execution, or other concrete malicious behavior.
Decision evidence
public snapshot- dist/index.js has explicit enable/login commands that write Claude Code settings and Codex shell profile wrappers.
- dist/index.js writes ~/.claude/settings.json statusLine/spinnerTipsOverride and backups under ~/.claude.
- dist/index.js writes marked codex() shell functions into ~/.zshrc, ~/.bashrc, fish, or PowerShell profiles.
- dist/adapters/* read ~/.platfory/auth.json and send Bearer-authenticated ad/impression requests.
- dist/index.js update command runs npm install -g @platfory/cli@latest when user invokes platfory update.
- package.json defines only a bin entry and no preinstall/install/postinstall lifecycle hooks.
- AI tool config mutation is reached via user-invoked login/enable with confirmation unless --yes is provided.
- README.md describes the sponsored AI-tool integrations and says changes are consent-gated and reversible.
- Restore/disable paths remove the marked Codex block and restore or clean Claude settings.
- Source sanitizes ad text/control bytes and validates https click URLs in reviewed paths.
- No code/prompt file harvesting, broad filesystem collection, remote payload execution, or credential exfiltration found.
Source & flagged code
5 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/index.jsView on unpkgPackage source invokes a package manager install command at runtime.
dist/index.jsView on unpkg · L868Source writes installer persistence such as shell profile or service configuration.
dist/index.jsView on unpkg · L3