OpenClaw Studio
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This looks like a local OpenClaw dashboard, but it includes persistent macOS background-service installers for auto-heal/watchdog code that is not included in the reviewed bundle.
Use the simple foreground run path first, and do not run install_launchd.sh until the missing server, auto-heal, watchdog, and dependency files are present and reviewed. If you do enable launchd services, confirm what each helper does, verify the uninstall script works, and limit config.json to only the local OpenClaw folders you intend to monitor.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If the user runs this installer, background services may keep starting at login and continue monitoring or healing the local OpenClaw agent until explicitly removed.
The installer creates and immediately starts persistent LaunchAgents for the monitor, auto-heal, and watchdog services. That is materially broader than a one-time local dashboard run.
<key>RunAtLoad</key><true/> ... <key>KeepAlive</key><true/> ... <key>StartInterval</key><integer>30</integer> ... launchctl bootstrap ... launchctl kickstart
Only run the launchd installer if persistent background monitoring is desired. Prefer manual run mode first, and make monitor, auto-heal, and watchdog separate explicit opt-ins with clear uninstall steps.
The reviewed bundle cannot prove what the auto-heal/watchdog services would do, and the installer may fail or later run code that was not part of this review.
The persistent services launch Python files that are not present in the supplied file manifest. Other referenced files such as requirements.txt, index.html, and monitor_config.py are also absent, leaving key runtime behavior unreviewable.
<string>$DIR/server.py</string> ... <string>$DIR/autoheal.py</string> ... <string>$DIR/app_watchdog.py</string>
Include the complete runtime source, dependency files, and lockfiles in the skill package, or remove the installer until those files are available and reviewable.
The dashboard or helpers may display or act on local OpenClaw state and case data from the configured folders.
The default configuration points the dashboard at local OpenClaw state and case folders. That is purpose-aligned for monitoring, but it may involve local session or case context.
"home": "~/.openclaw", "session_key": "agent:main:main", "case_roots": ["~/Documents/OpenClaw/cases"]
Review config.json before use, keep paths narrow, avoid pointing it at unrelated private folders, and keep cloud sharing disabled unless explicitly intended.