@luckydraw/cumulus@0.31.66
RLM-based CLI chat wrapper for Claude with external history context management
AI Security Review
scanned 3mo ago · by lpm-firewall-aiSource inspection found powerful but package-aligned, user-invoked agent and gateway functionality, not a concrete malicious chain. Install-time hooks do not perform runtime mutation.
Decision evidence
public snapshot- Agent tools can run shell commands during an explicit Claude/gateway session (dist/lib/tools/bash.js).
- package.json has no preinstall/install/postinstall hook.
- Gateway persistence is only offered after explicit cumulus-gateway setup confirmation (dist/gateway/setup.js).
- Runtime update installs are reachable through explicit update commands or authenticated admin API (dist/lib/version-check.js, dist/gateway/server.js).
- API-key-based authentication gates non-public gateway routes (dist/gateway/auth.js, dist/gateway/server.js).
- No credential-harvesting or covert exfiltration path was found in inspected source.
Source & flagged code
6 flagged · loading sourcePackage source references child process execution.
dist/tui/hooks/useClaudeProcess.jsView on unpkg · L3Findings
Affected versions and remediation
This report applies to @luckydraw/cumulus@0.31.66.
See all versions of @luckydraw/cumulus for other recorded verdicts.
- Review the evidence and your use of @luckydraw/cumulus@0.31.66 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.