Browser Healthcheck
v1.0.0Automatically checks and repairs browser tool status before use to prevent timeouts, CDP disconnections, and profile conflicts during automation.
⭐ 0· 55·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (browser health check / auto-repair) align with the included script and SKILL.md. The script inspects a local OpenClaw config (~/.openclaw/openclaw.json), probes localhost CDP ports, and offers to start/kill/restart services — all coherent with diagnosing browser/CDP issues.
Instruction Scope
Instructions stay focused on browser health: checking status, querying http://127.0.0.1:<port>/json endpoints, and restarting or killing local processes. Two items to note: (1) the SKILL.md example uses exec("openclaw gateway restart") inside Python, which is a coding error (exec executes Python, not shell) — this is a bug, not exfiltration; (2) the post-fix actions include running system commands (taskkill, openclaw gateway restart) which are expected for repair but are privileged actions the user should expect before enabling automatic fixes.
Install Mechanism
There is no install spec (instruction-only plus a single Python script). Nothing is downloaded or written by an installer during skill install.
Credentials
The skill requests no environment variables, no external credentials, and only reads a local configuration file (~/.openclaw/openclaw.json). That local config access is reasonable for a tool that needs to inspect OpenClaw browser settings.
Persistence & Privilege
The skill is not always-enabled and uses the platform defaults for invocation. The only privileged actions are local repair steps (taskkill, restarting gateway) which are gated behind the --fix flag or explicit repair instructions — appropriate for this purpose.
Assessment
This skill appears to do what it says: local diagnostics of the OpenClaw browser/CDP setup. Before installing or running with automatic fixes enabled, consider: (1) The script reads ~/.openclaw/openclaw.json — review that file if you care about what configuration is inspected. (2) If you run the script with --fix or follow SKILL.md repair steps, it may run system commands that kill processes (taskkill) or restart the OpenClaw gateway — these are destructive to local processes and require appropriate privileges. (3) Some repair commands in SKILL.md are Windows-specific (netstat -ano, taskkill) and the find/kill logic may not work on Linux/macOS as-is. (4) The SKILL.md contains a small coding mistake (using exec("openclaw gateway restart") in a Python snippet) — the intent is to run the shell command, but that snippet will not work as shown. Recommendation: run the health check in non-fix/diagnostic mode first to see reported issues, inspect the detected PID/process before allowing automated kills, and ensure the skill is only used on machines where you trust restarting/killing processes is acceptable.Like a lobster shell, security has layers — review code before you run it.
latestvk972x69wx20ywqdxq100zwgrvd83sn6q
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
