@luckydraw/cumulus@1.0.41
RLM-based CLI chat wrapper for Claude with external history context management
AI Security Review
scanned 3d ago · by lpm-firewall-aiSource inspection confirms that the package launches Claude with permission prompts bypassed, including in the TUI, while gateway restrictions depend on operator configuration. This supports a warning for dangerous agent capability, but does not establish a separate malicious attack or install-time behavior.
Decision evidence
public snapshot- The package exposes a Claude wrapper and gateway; both gateway and TUI launch Claude with permission prompts bypassed.
- Gateway deny-by-default restrictions depend on an allowlist being configured; without one, the existing denylist is retained, and tool access is otherwise available.
- The gateway also accepts turns through configured chat, webhook, scheduled, and API paths. API routes include key and thread-scope checks; these controls do not restore per-action Claude approval.
- Inspected flagged runtime locations concern formatting, file summaries, task handling, and configured agent messaging; they do not establish unrelated account actions or another attack.
- The manifest identifies the package as a Claude wrapper and uses build and Husky setup lifecycle scripts; no install-time attack was established.
- Gateway tool restrictions can be configured through allowed and disallowed tool settings.
Source & flagged code
8 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@1.0.41.
See all versions of @luckydraw/cumulus for other recorded verdicts.
- Review the evidence and your use of @luckydraw/cumulus@1.0.41 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.