thalamus@1.5.18
Mobile and Web client for Claude Code
AI Security Review
scanned 1mo ago · by lpm-firewall-aiInspected install behavior stays within package tool setup, and the flagged agent paths support session lookup rather than configuration mutation. Runtime process control and authentication flows align with the client functionality, with no concrete malicious chain identified.
Decision evidence
public snapshot- The postinstall hook extracts bundled platform tools into the package tools directory; no runtime self-dependency is declared.
- Both flagged API locations resolve Claude project session paths for session checks, without modifying agent configuration.
- The MCP launcher starts the package bridge and forwards caller arguments and termination signals.
- The CLI supplies its own MCP bridge through session launch arguments; watchdog installation is an explicit command.
- The inspected difftastic archive contains a difft executable and is unpacked as a local comparison tool.
- Both library entrypoints expose API clients and configuration from the inspected API bundles.
- Bundled native executable internals were not fully analyzed; the CLI supports remote session control as part of its functionality.
Source & flagged code
14 flaggedPackage defines install-time lifecycle scripts.
package.jsonView on unpkgPackage source references child process execution.
bin/thalamus-mcp.mjsView on unpkg · L2Findings
Affected versions and remediation
This report applies to thalamus@1.5.18.
See all versions of thalamus for other recorded verdicts.
- Review the evidence and your use of thalamus@1.5.18 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.