| name | run-coverage |
| description | Run the Budget Buddy test suite with a coverage report. Use when asked to run tests, check coverage, see the coverage report, or verify the coverage gate before or after making changes. |
Skill: Run Test Coverage
This skill runs the Budget Buddy pytest suite and prints a coverage report so you
can see which lines are not yet tested.
How to run it
Run the script for the current platform from this skill's directory:
- On Linux or macOS (including the GitHub Copilot cloud agent), run
run-coverage.sh.
- On Windows (local VS Code agent mode), run
run-coverage.ps1.
Each script:
- Selects the project virtual environment (
.venv) if present, otherwise falls
back to the system python.
- Runs
pytest with --cov=. --cov-report=term-missing.
Note: this skill does not pre-approve the shell tool, so Copilot will ask for
confirmation before running the script. To skip that prompt in a trusted demo,
add allowed-tools: shell to the frontmatter above. Only do this if you trust
the skill and its scripts.
Interpreting the result
- The
Missing column lists line numbers with no test coverage. These are good
candidates for new tests.
- The run fails if total coverage is below the gate in
pytest.ini. During the
demo this gate is raised from 30 to 90.
After running
Summarize the current coverage percentage and call out the modules with the most
uncovered lines.
