claude-in-chrome-troubleshooting

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides numerous shell commands for manipulating Chrome's Native Messaging Host configuration files (JSON) in ~/Library/Application Support/Google/Chrome/NativeMessagingHosts/ to resolve conflicts between different versions of the Claude bridge.
  • [COMMAND_EXECUTION]: Instructs the agent to create a local bash wrapper script at ~/.claude/chrome/chrome-native-host using shell redirection (cat > ...) and permission modification (chmod +x). The script dynamically identifies the latest version of Claude Code to execute.
  • [COMMAND_EXECUTION]: Includes diagnostic procedures that utilize system tools such as ps, pgrep, pkill, lsof, and osascript to manage local browser automation processes and verify socket connectivity.
  • [COMMAND_EXECUTION]: Suggests appending a utility function and environment variable exports (e.g., TMPDIR) to the user's shell profile (~/.zshrc) to assist in ongoing maintenance and debugging.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 10:19 PM
Security Audit — agent-trust-hub — claude-in-chrome-troubleshooting