@aggroot-team/aggroot@1.12.1
Package no longer supported. Contact Support at https://www.npmjs.com/support for more info.
AggRoot - AI Assistant (Node.js/TypeScript Version)
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThe package is a user-invoked AI assistant CLI with no npm lifecycle hook and no hidden install chain. Dual-use shell, SSH, and browser features stay behind that explicit CLI and match the product, so they do not form a concrete publish-blocking attack.
Decision evidence
public snapshot- package.json exposes the aggroot bin at dist/cli.cjs and its scripts object has no preinstall, install, or postinstall hook.
- dist/cli.cjs only re-spawns the current Node binary with a larger heap and then loads dist/index.cjs.
- scripts/install-browsers.mjs runs npx playwright install chromium against npmmirror or the official host, and it is not registered as a lifecycle script.
- scripts/test-tokenizer.js is a local shell-lexer helper with no network or install side effects.
- The cited share/web assets are Mermaid state and flowchart diagram parsers.
- Self-update uses registry.npmjs.org and npm install -g only from the user upgrade command, not from install.
- The bundled assistant can run shells, SSH, and browser tools after the user starts the CLI.
- javascript-obfuscator is listed only as a devDependency.
- install-linux.sh is a manual installer and is not an npm lifecycle hook.
Source & flagged code
12 flaggedPackage source references dynamic require/import behavior.
dist/cli.cjsView on unpkg · L2Package source references shell execution.
scripts/install-browsers.mjsFindings
Affected versions and remediation
This report applies to @aggroot-team/aggroot@1.12.1.
See all versions of @aggroot-team/aggroot for other recorded verdicts.
- Review the evidence and your use of @aggroot-team/aggroot@1.12.1 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.