What problem does it solve?
This Skill eliminates the risk of inconsistent, insecure, or non-interoperable AIASys Agent implementations by providing standardized development guidelines for tools, session management, sandbox execution, and streaming output.
Core Features & Use Cases
- Tool Development Standards: Provides a 5-element implementation template covering complete docstrings, parameter validation, sandbox execution, timeout control, and exception handling for all Agent tools.
- Cross-Platform Shell Support: Includes guidelines for handling Windows/POSIX shell differences, interpreter fallback chains, and command escaping to ensure tools work across operating systems.
- Agent State Management: Defines strict state machine patterns for session lifecycle management, including state persistence, breakpoint resumption, and task cancellation for long-running Agent tasks.
- Sandbox Security & Streaming: Covers sandbox resource limits, security principles, and SSE streaming implementation for real-time, secure Agent execution feedback.
- Use Case: For example, when building a new data analysis Agent tool for AIASys, use this Skill to ensure the tool follows parameter validation rules, executes code in a sandbox with proper timeouts, and integrates with the platform's session state management for resumable long-running tasks.
Quick Start
Use the aiasys-tool-dev skill to create a new compliant Agent tool for executing Python data analysis code in a sandboxed environment with proper timeout and error handling.
