@luckydraw/cumulus@0.31.54
RLM-based CLI chat wrapper for Claude with external history context management
AI Security Review
scanned 3mo ago · by lpm-firewall-aiThe source establishes a concrete dangerous AI-agent capability and embeds prompt-level control instructions, so it should be warned. It does not meet the block threshold because no unconsented install hook, foreign AI-control-surface mutation, or autonomous exfiltration chain was found.
Decision evidence
public snapshot- dist/lib/gateway.js embeds extensive AI-agent workflow/tool-routing instructions, including commands to suppress normal tooling.
- dist/mcp/tool-handler.js resolves and reads any supplied absolute file path, then persistently stores its contents.
- dist/lib/gateway.js launches Claude with --permission-mode bypassPermissions.
- dist/lib/tools/bash.js exposes bash -c execution to the agentic tool set.
- dist/lib/tools/web-fetch.js permits arbitrary HTTP methods and URLs from agent tool input.
- package.json has no preinstall, install, or postinstall hook; prepare only invokes husky.
- Gateway service persistence in dist/gateway/setup.js occurs only after explicit cumulus-gateway setup and interactive consent.
- Runtime npm installation in dist/lib/version-check.js is reached only by explicit update/rollback commands.
- Observed fixed external endpoints serve declared package functions; no covert credential-exfiltration endpoint was found.
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@0.31.54.
See all versions of @luckydraw/cumulus for other recorded verdicts.
- Review the evidence and your use of @luckydraw/cumulus@0.31.54 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.