- Create new AgentExecutionService.ts with secure agent.py script execution
- Replace arbitrary shell command execution with controlled Python script calls
- Add claude_session_id field to session types for conversation continuity
- Update shared types between main and renderer processes
- Implement proper argument validation and sanitization
- Add comprehensive error handling and logging
- Export service through agent service index
Security improvements:
- Only executes predefined agent.py script (no arbitrary commands)
- Uses direct process spawning instead of shell execution
- Validates all arguments before execution
- Prevents command injection vulnerabilities
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Vaayne