08-error-handling.yaml 1020 B

123456789101112131415161718192021222324252627282930313233343536373839404142434445
  1. id: golden-08-error-handling
  2. name: "Golden 08: Error Handling - Agent Handles Missing Files Gracefully"
  3. description: |
  4. Tests that the agent handles errors gracefully when files don't exist.
  5. A quality agent should:
  6. 1. Attempt to read the requested file
  7. 2. Receive an error (file not found)
  8. 3. Report the error clearly to the user
  9. 4. Not crash or behave unexpectedly
  10. This is a read-only test that checks error handling behavior.
  11. Validates:
  12. - Agent attempts the read operation
  13. - Agent handles the error gracefully
  14. - Agent communicates the issue to the user
  15. category: developer
  16. prompts:
  17. - text: |
  18. Read the file evals/test_tmp/this-file-does-not-exist-12345.txt and show me its contents.
  19. approvalStrategy:
  20. type: auto-approve
  21. behavior:
  22. mustUseTools:
  23. - read
  24. minToolCalls: 1
  25. maxToolCalls: 3
  26. expectedViolations:
  27. - rule: approval-gate
  28. shouldViolate: false
  29. severity: error
  30. timeout: 60000
  31. tags:
  32. - golden
  33. - error-handling
  34. - missing-file
  35. - read-only
  36. - safe