@eventmodelers/cli@1.0.55
Eventmodelers CLI — real-time Claude agent + skills for Claude Code, for any stack (Node, Supabase, Axon, Cratis, OpenCQRS, UmaDB, Kurrent, or modeling-only)
AI Security Review
scanned 25d ago · by lpm-firewall-aiThe high-risk indicators point to explicit CLI setup and project tooling, while the manifest has no lifecycle scripts. The inspected sources do not establish an active attack or unconsented install-time agent-control mutation.
Decision evidence
public snapshot- package.json exposes cli.js as the command entrypoint and defines no lifecycle scripts.
- cli.js maps agent-host integrations to command or workflow directories; the inspected code writes these files through the explicit init-agents command.
- The bridge Claude runner uses Claude Code's skip-permissions option when processing queued work.
- The bridge shell hook requires a configured command before running it.
- The exporter serves JSON over HTTP; the inspected excerpt does not establish credential exfiltration.
- The TypeScript check runs the project's own npx tsc command; the adapter imports child-process and file helpers and sets the Eventmodelers API as its default endpoint.
- The package manifest has no install-time lifecycle hook, so the inspected CLI configuration and project checks are not automatically run during npm installation.
Source & flagged code
9 flaggedPackage source references child process execution.
stacks/bridge/templates/bridge/ralph-claude.jsView on unpkg · L8Package source references shell execution.
stacks/bridge/templates/bridge/ralph-hook.jsView on unpkg · L68Findings
Affected versions and remediation
This report applies to @eventmodelers/cli@1.0.55.
See all versions of @eventmodelers/cli for other recorded verdicts.
- Review the evidence and your use of @eventmodelers/cli@1.0.55 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.