slaunt@0.1.8
Privacy-first MCP agent access auditor for Claude and Codex
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package has a real, explicit-user-command AI-agent configuration mutation capability. It lacks install hooks and no concrete stealth, exfiltration, payload, or persistence behavior was found, so this warrants a warning rather than a block.
Decision evidence
public snapshot- The explicit --install-slaunt command invokes an installer that changes detected AI-client configuration files.
- The installer adds a Slaunt MCP entry to Claude and Codex configuration files, using a remote HTTPS endpoint.
- The CLI can also start configured local MCP servers only after approval or an explicit automation flag.
- package.json has no preinstall, install, or postinstall hook.
- Configuration mutation requires the user to pass --install-slaunt; it is not automatic at install time.
- The installer refuses symlinks, backs up existing files, and limits its own endpoint to HTTPS.
- No credential harvesting, secret exfiltration, hidden payload execution, or self-dependency was found.
Source & flagged code
4 flaggedPackage metadata claims a different repository identity while copied source loads a runtime dependency bridge.
dist/explanation/local.jsView on unpkg · L233A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/explanation/local.js#virtual:normalized:round1View on unpkgFindings
Affected versions and remediation
This report applies to slaunt@0.1.8.
See all versions of slaunt for other recorded verdicts.
- Review the evidence and your use of slaunt@0.1.8 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.