@luckydraw/cumulus@0.31.64
RLM-based CLI chat wrapper for Claude with external history context management
AI Security Review
scanned 3mo ago · by lpm-firewall-aiThe package is not concretely malicious, but it exposes high-impact agentic shell and network capabilities without per-tool confirmation. Its persistence feature is explicit and package-owned.
Decision evidence
public snapshot- dist/lib/tools/index.js exposes Bash and WebFetch to every agentic turn.
- dist/lib/tools/bash.js executes model-provided shell commands in the selected cwd.
- dist/lib/tools/web-fetch.js permits arbitrary HTTP methods, URLs, headers, and bodies.
- dist/gateway/setup.js can install persistent LaunchAgent/systemd services after an explicit setup prompt.
- package.json has no preinstall/install/postinstall hook; prepare only invokes husky.
- Service setup is reached only through explicit cumulus-gateway setup and asks for confirmation.
- No source evidence of credential harvesting, covert exfiltration, obfuscated payloads, or foreign AI-agent config writes.
Source & flagged code
7 flagged · loading sourcePackage source references child process execution.
dist/tui/hooks/useClaudeProcess.jsView on unpkg · L3This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/lib/gateway.jsView on unpkgFindings
Affected versions and remediation
This report applies to @luckydraw/cumulus@0.31.64.
See all versions of @luckydraw/cumulus for other recorded verdicts.
- Review the evidence and your use of @luckydraw/cumulus@0.31.64 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.