终端命令行Agent工具链集成测试生成器

You are a senior QA engineer specializing in CLI-based AI agent testing. Generate comprehensive integration test suites for agent tool chains.

Agent Tool Chain Description: [Paste your agent tools/skills configuration here]

Generate test cases covering:

1. Happy Path Tests: Standard tool invocation sequences that should succeed
2. Permission Tests: Verify tools respect authorization boundaries
3. Input Validation: Edge cases for each tool parameter (empty, oversized, special chars, unicode)
4. Timeout & Latency: Simulate slow responses and verify timeout handling
5. Dependency Failures: What happens when external services (APIs, DBs, file systems) are unavailable
6. Concurrency Tests: Multiple simultaneous tool calls, race conditions
7. State Management: Tool calls that depend on previous results, context window limits
8. Rollback Scenarios: Verify cleanup when multi-step operations fail midway

For each test case provide:

• Test name and description
• Setup/preconditions
• Input command or sequence
• Expected output/behavior
• Cleanup steps
• Priority (P0-P3)

Output format: structured test plan with shell script snippets ready to execute.