registry / @eventmodelers/cli / 2.0.44
@eventmodelers/cli@2.0.44
Eventmodelers CLI — real-time Claude agent + skills for Claude Code, for any stack (Node, Supabase, Axon, OpenCQRS, UmaDB, Kurrent, or modeling-only)
AI Security Review
scanned 1d ago · by lpm-firewall-aiThe manifest has no install hooks, and the flagged credential and agent-configuration behavior is consistent with a CLI that runs an agent against a configured provider and sets up project skills on request. The reviewed code does not establish automatic forwarding of credentials to a fixed unofficial host.
Static reason
One or more suspicious static signals were detected.
Trigger
The CLI runs when invoked; prototype project changes occur through an explicit init --prototype command.
Impact
No automatic credential redirection or unrequested account action was established from the inspected source.
Mechanism
The agent runner forwards a configured provider base URL, and prototype setup writes project-local skills and permissions that deny board write tools.
AI rationale
The manifest has no install hooks, and the flagged credential and agent-configuration behavior is consistent with a CLI that runs an agent against a configured provider and sets up project skills on request. The reviewed code does not establish automatic forwarding of credentials to a fixed unofficial host.
Evidence
.claude/settings.local.jsonCLAUDE.md.claude/skills/
Network endpoints2
api.eventmodelers.aiapi.franconia.ai/v1
Decision evidence
public snapshotAI called this Clean at 89.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes a CLI and declares no install, preinstall, or postinstall lifecycle hook.
- The CLI passes ANTHROPIC_BASE_URL only when configured; the inspected code does not set it to a fixed unofficial gateway.
- The example Franconia endpoint is documented as an optional Mastra endpoint, with its key read from MASTRA_API_KEY.
- Prototype setup is reached through the explicit init --prototype option.
- Prototype setup writes project-local skills and limits its settings change to denying board write tools.
Evidence against
- The CLI can pass a configured provider endpoint to an agent process, so the user must choose that endpoint.
- Explicit prototype initialization modifies project files, including Claude skills, CLAUDE.md, and local permissions settings.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
Source & flagged code
11 flaggedstacks/supabase/templates/root/src/common/testHelpers.tsView file
1import {execSync} from 'child_process';
L2: import {readFileSync, writeFileSync, unlinkSync} from 'fs';
High
Child Process
Package source references child process execution.
stacks/supabase/templates/root/src/common/testHelpers.tsView on unpkg · L1shared/build-kit/ralph-exec.jsView file
121stdio: ['inherit', traced ? 'pipe' : 'inherit', 'inherit'],
L122: shell: true,
L123: env: { ...childEnv, RALPH_PROMPT_FILE: promptFile, ...extraEnv },
High
stacks/supabase/templates/root/server.tsView file
Findings
8 High5 Medium5 Low
HighChild Processstacks/supabase/templates/root/src/common/testHelpers.ts
HighShellshared/build-kit/ralph-exec.js
HighSame File Env Network Executionshared/build-kit/code-export.mjs
HighCredential Redirect Persistencecli.js
HighExternal Ai Agent Control Surface Mutationcli.js
HighTrigger Reachable External Ai Agent Control Surface Mutationextensions/prototype/index.js
HighRuntime Package Installstacks/supabase/templates/build-kit/lib/checks/90-tsc-build.cjs
HighSemantic Analysis Limitedlib/adapters/spec-kitty-adapter.js#virtual:normalized:round1
MediumDynamic Requirestacks/supabase/templates/root/server.ts
MediumNetwork
MediumEnvironment Vars
Affected versions and remediation
This report applies to @eventmodelers/cli@2.0.44.
See all versions of @eventmodelers/cli for other recorded verdicts.
- Review the evidence and your use of @eventmodelers/cli@2.0.44 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.