@roarkanalytics/sdk-mcp@4.24.0
The official MCP Server for the Roark API
AI Security Review
scanned 3d ago · by lpm-firewall-aiThe package does not show a concrete malicious action or install-time behavior, but its default agent-invoked code runner exposes the parent process environment to submitted code. This warrants a warning for dangerous local code execution capability.
Decision evidence
public snapshot- The manifest points to index.js and contains no install, preinstall, or postinstall hook.
- The server includes the execute tool by default, and the CLI defaults code execution to local mode.
- The local worker receives submitted code and evaluates it with the authenticated SDK client.
- The Deno worker receives the parent process environment and is granted environment access.
- The worker's network permission is limited to the configured Roark API hostname.
- Code execution requires an execute tool call with supplied code; no automatic execution at installation was identified.
- The inspected local worker network grant is scoped to the configured API hostname.
Source & flagged code
12 flagged · loading sourceSource file is highly similar to a previously finalized malicious package; route for source-aware review.
instructions.jsView on unpkgFindings
Affected versions and remediation
This report applies to @roarkanalytics/sdk-mcp@4.24.0.
See all versions of @roarkanalytics/sdk-mcp for other recorded verdicts.
- Review the evidence and your use of @roarkanalytics/sdk-mcp@4.24.0 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.