agenticros@0.7.23
AgenticROS - agentic AI for ROS-powered robots. Single CLI to launch real-robot or simulated demos, manage configuration, and inspect status.
AI Security Review
scanned 27d ago · by lpm-firewall-aiThe package is a robot CLI that changes Claude, Codex, and OpenClaw configuration only after an explicit command, with prompts for plugin install and MCP registration. The alsoAllow expansion during init and the opt-in gateway auth-none script are guarded dual-use agent control changes, not an install-time hijack.
Decision evidence
public snapshot- package.json publishes a CLI bin and has no preinstall, install, or postinstall hook.
- agenticros init asks before installing the OpenClaw plugin and before writing Codex, Hermes, and Claude MCP config.
- init then calls syncToolsAlsoAllow, which appends plugin tool ids to tools.alsoAllow in the user OpenClaw config when that file and the plugin manifest already exist.
- Explicit mcp setup writes an agenticros stdio server entry into Claude Desktop or project MCP config and the matching Codex config.
- runtime/scripts/setup-openclaw-local.cjs sets gateway.auth.mode to none and deletes the gateway token when the user runs that script.
- Agent config writes are not triggered by npm install lifecycle scripts.
- MCP setup only replaces the agenticros server entry and points it at the package MCP program.
- The auth-none script is a separate opt-in local helper and is not referenced from package.json scripts.
Source & flagged code
23 flagged · loading sourcePackage source references child process execution.
dist/util/mcp-discovery.jsView on unpkg · L16Source file is highly similar to a previously finalized malicious package; route for source-aware review.
dist/util/mcp-discovery.jsView on unpkgFindings
Affected versions and remediation
This report applies to agenticros@0.7.23.
See all versions of agenticros for other recorded verdicts.
- Review the evidence and your use of agenticros@0.7.23 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.