|
| 1 | +# write-fastdeploy-unittest |
| 2 | + |
| 3 | +A skill that guides AI agents to write CI-compliant unit tests for the FastDeploy project. |
| 4 | + |
| 5 | +## Features |
| 6 | + |
| 7 | +- Automatically selects the appropriate test pattern based on the code under test (pure logic / GPU kernel / offline inference / E2E serving) |
| 8 | +- Follows FastDeploy CI classification rules (multi-GPU sequential vs single-GPU parallel) |
| 9 | +- Meets the 80% diff coverage PR threshold |
| 10 | +- Correctly uses port variables, log isolation, and resource cleanup per CI conventions |
| 11 | + |
| 12 | +## Usage |
| 13 | + |
| 14 | +### Basic — specify a source file |
| 15 | + |
| 16 | +``` |
| 17 | +Use the write-fastdeploy-unittest skill to add unit tests for fastdeploy/cache_manager/transfer_factory/file_store/file_store.py |
| 18 | +``` |
| 19 | + |
| 20 | +### From coverage report — paste the line directly |
| 21 | + |
| 22 | +``` |
| 23 | +Use the write-fastdeploy-unittest skill to add unit tests for: |
| 24 | +
|
| 25 | +fastdeploy/model_executor/model_loader/default_loader.py 48 32 14 0 26% 37-38, 42, 46-52, 56-66, 69-97 |
| 26 | +``` |
| 27 | + |
| 28 | +The coverage report format is: `file_path Stmts Miss Branch BrMiss Cover% Missing_lines`. The agent will focus on the uncovered lines and write tests specifically targeting those branches. |
| 29 | + |
| 30 | +The agent will automatically: |
| 31 | +1. Read the target source file and analyze uncovered lines |
| 32 | +2. Select the appropriate test pattern (Pattern 1-4) |
| 33 | +3. Generate a test file in the corresponding `tests/` subdirectory |
| 34 | +4. Run tests and verify coverage |
| 35 | + |
| 36 | +## Test Pattern Quick Reference |
| 37 | + |
| 38 | +| Pattern | Use Case | Dependencies | |
| 39 | +|---------|----------|--------------| |
| 40 | +| 1 — Pure Logic | config, utils, scheduler, router, etc. | No GPU; mock external deps | |
| 41 | +| 2 — GPU Kernel | ops, layers, numerical computation | Requires GPU; `@pytest.mark.gpu` | |
| 42 | +| 3 — Offline Inference | LLM API, model loading | Requires MODEL_PATH | |
| 43 | +| 4 — E2E Serving | End-to-end HTTP serving | subprocess + ports | |
| 44 | + |
| 45 | +## Key Conventions |
| 46 | + |
| 47 | +- Test file naming: `test_<module>.py` |
| 48 | +- Test class naming: `Test<Module>` |
| 49 | +- Coverage verification: `python -m coverage run --source=<directory> -m pytest <test_file> && coverage report -m` |
| 50 | +- The `--source` parameter must be a directory path, not a dotted module name |
| 51 | + |
| 52 | +## Related Files |
| 53 | + |
| 54 | +- [SKILL.md](SKILL.md) — Full skill instruction document |
0 commit comments