John Smith d0d1afc383 feat(harbor): add Harbor Task Format adapter with 3 sample tasks ...

- benchmarks/harbor_adapter.py: HarborBenchmark loads tasks from directories
  containing instruction.md + test.sh + optional Dockerfile/verify_files.json
- benchmarks/__init__.py: register 'harbor' in resolve_benchmark with
  configurable dataset_path (defaults to tasks/)
- tasks/: 3 sample Harbor tasks (add-numbers, create-file, sort-list)
  with instruction, test script, and file-state assertions
- tests/test_harbor.py: 11 tests covering parsing, discovery, grading,
  file-state checks, and edge cases (147 total tests now pass)

2026-08-22 13:30:23 -04:00

..

add-numbers

feat(harbor): add Harbor Task Format adapter with 3 sample tasks

2026-08-22 13:30:23 -04:00

create-file

feat(harbor): add Harbor Task Format adapter with 3 sample tasks

2026-08-22 13:30:23 -04:00

sort-list

feat(harbor): add Harbor Task Format adapter with 3 sample tasks

2026-08-22 13:30:23 -04:00