uv-pytest-unit-testing（uv pytest 单元测试）

Uv Pytest Unit Testing

Purpose

Use this skill to standardize pytest setup and execution for uv-managed Python repositories, including single-project repos and uv workspaces.

When To Use

• Use this skill for pytest setup, execution, and troubleshooting in uv-managed repositories.
• Use this skill when the user wants package-targeted pytest runs for a workspace member.
• Keep scope on test setup and execution; do not use this as a generic repo bootstrap skill.

Primary Workflow

1. Detect repository mode.

- Treat the repo as a workspace when pyproject.toml defines [tool.uv.workspace]. - Treat it as a single project otherwise.

1. Bootstrap pytest dependencies and baseline config:

- scripts/bootstrap_pytest_uv.sh --workspace-root <repo> - add --package <member-name> for workspace-member setup - add --with-cov when the task explicitly wants pytest-cov - add --dry-run when previewing changes

1. Run tests:

- scripts/run_pytest_uv.sh --workspace-root <repo> - scripts/run_pytest_uv.sh --workspace-root <repo> --package <member-name>

1. Troubleshoot failures in this order:

- command context - test discovery layout - marker registration - import-path assumptions

Test Authoring Guidance

• Keep fast unit tests under tests/unit and integration-heavy tests under tests/integration when repo size warrants separation.
• Use fixtures for setup reuse, and keep fixture scope minimal (function by default).
• Use @pytest.mark.parametrize for matrix-style cases instead of hand-written loops.
• Use monkeypatch for environment variables and runtime dependency replacement.
• When generated projects use pydantic-settings, prefer overriding environment variables or the settings dependency rather than mutating committed .env files in tests.
• Register custom marks in config to avoid marker warnings.

Automation Suitability

• Codex App automation: High. Strong recurring fit for test health checks, failure triage, and drift detection.
• Codex CLI automation: High. Strong fit for non-interactive test setup and targeted test sweeps.

Codex App Automation Prompt Template

Use $uv-pytest-unit-testing.

Scope boundaries:
- Work only inside <REPO_PATH>.
- Operate only on pytest setup and test execution tasks.
- Do not perform unrelated code refactors.

Task:
1. Detect repository mode from <WORKSPACE_ROOT>/pyproject.toml.
2. If <DRY_RUN_BOOTSTRAP:TRUE|FALSE> is TRUE, run:
   `scripts/bootstrap_pytest_uv.sh --workspace-root <WORKSPACE_ROOT> <PACKAGE_FLAG> <WITH_COV_FLAG> --dry-run`
3. If <DRY_RUN_BOOTSTRAP:TRUE|FALSE> is FALSE, run:
   `scripts/bootstrap_pytest_uv.sh --workspace-root <WORKSPACE_ROOT> <PACKAGE_FLAG> <WITH_COV_FLAG>`
4. Run tests with:
   `scripts/run_pytest_uv.sh --workspace-root <WORKSPACE_ROOT> <PACKAGE_FLAG> <TEST_PATH_FLAG> -- <PYTEST_ARGS>`
5. Keep package-targeted runs explicit when <PACKAGE_NAME_OR_EMPTY> is set.

Output contract:
1. STATUS: PASS or FAIL
2. SETUP: what bootstrap actions ran
3. TEST_RESULTS: concise pass/fail summary
4. FAILURES: grouped likely causes
5. NEXT_STEPS: minimal remediation actions

Codex CLI Automation Prompt Template

codex exec --full-auto --sandbox workspace-write --cd "<REPO_PATH>" "<PROMPT_BODY>"

<PROMPT_BODY> template:

Use $uv-pytest-unit-testing.
Limit scope to pytest setup and execution in <WORKSPACE_ROOT>.
Run bootstrap in dry-run or real mode based on <DRY_RUN_BOOTSTRAP:TRUE|FALSE>.
Run tests with explicit package targeting when <PACKAGE_NAME_OR_EMPTY> is set.
Return STATUS, setup actions, concise test summary, grouped likely causes for failures, and minimal next steps.

Customization Placeholders

• <REPO_PATH>
• <WORKSPACE_ROOT>
• <PACKAGE_NAME_OR_EMPTY>
• <PACKAGE_FLAG>
• <TEST_PATH_OR_EMPTY>
• <TEST_PATH_FLAG>
• <PYTEST_ARGS>
• <WITH_COV_FLAG:--with-cov|EMPTY>
• <DRY_RUN_BOOTSTRAP:TRUE|FALSE>

Interactive Customization Workflow

1. Ask whether users want bootstrap mode or run mode.
2. Gather workspace root and optional package target.
3. For bootstrap mode, gather with_cov and dry_run.
4. For run mode, gather optional test path and optional pytest args.
5. Return both:

• A YAML profile for durable reuse.
• The exact command to run.

1. Use this precedence order:

• CLI flags
• --config profile file
• .codex/profiles/uv-pytest-unit-testing/customization.yaml
• ~/.config/gaelic-ghost/python-skills/uv-pytest-unit-testing/customization.yaml
• Script defaults

1. If users want temporary reset behavior:

• --bypassing-all-profiles
• --bypassing-repo-profile
• --deleting-repo-profile

1. If users provide no customization or profile files, keep existing script defaults unchanged.
2. See references/interactive-customization.md for schema and examples.

Outputs

• status

- success: setup or test execution completed - blocked: repo shape or prerequisites prevented the run - failed: bootstrap or test execution ran but did not complete cleanly

• path_type

- primary: canonical shell entrypoints completed

• output

- setup actions taken - exact run command used - concise failure grouping when applicable

Guardrails

• Require uv for all installation and execution paths.
• Require an existing pyproject.toml under --workspace-root.
• Keep package targeting explicit when operating on a workspace member.

References

• references/pytest-workflow.md
• references/uv-workspace-testing.md
• references/customization.md
• references/interactive-customization.md

Script Inventory

• scripts/bootstrap_pytest_uv.sh
• scripts/run_pytest_uv.sh