Three test-only tools with minimal descriptions and no input validation. 'echo' and 'fail' have trivial 10-char descriptions (below 20-char threshold). 'slow' lacks any description. All tools have basic schemas but lack parameter constraints, enums, or validation guidance. No error recovery hints. Tools are READ_ONLY and test-focused, not production-grade. Naming is generic (echo, fail, slow) rather than action-verb patterns (e.g., test_echo_string). No output schema documentation. This is a test harness, not a production tool suite.

• Expand 'echo' description from 'Echo back input text' to 'Echoes the input text back to the caller. Used to test tool invocation and parameter passing. Returns the exact text provided.' (60+ chars, actionable).
• Rename tools to action-verb patterns: 'echo' → 'test_echo_string', 'fail' → 'test_simulate_failure', 'slow' → 'test_long_running_operation'. This helps LLMs infer intent.
• Add output schema documentation. E.g., 'echo' returns {"echoed_text": string, "length": int}. 'fail' returns {"error_message": string, "error_code": string}. 'slow' returns {"duration_ms": int, "status": string}.
• Add parameter constraints to 'echo.text': min 1 char, max 10000 chars. Describe: 'The text to echo back (1-10000 characters). Used for testing parameter passing.'
• Add error recovery hint to 'fail': 'This tool intentionally fails for testing. If you see this error in production, the tool is misconfigured. Retry with different parameters or skip this step.'