Pytest Workflow

Pytest Workflow

Overview

Use this skill to design pytest suites that are fast enough for daily feedback and stable enough for CI gates.

Scope Boundaries

• Use this skill when the task matches the trigger condition described in description.
• Do not use this skill when the primary task falls outside this skill's domain.

Shared References

• Trigger examples:
  • references/trigger-and-examples.md
• Fixture boundary rules:
  • references/fixture-boundary-rules.md
• Determinism rules:
  • references/pytest-determinism-rules.md

Templates And Assets

• Test plan template:
  • assets/pytest-test-plan-template.md
• Fixture stability checklist:
  • assets/pytest-fixture-stability-checklist.md
• Command profile template:
  • assets/pytest-command-profile-template.md

Inputs To Gather

• Python modules and behaviors at risk.
• Fixture graph, dependency seams, and isolation constraints.
• Marker strategy and CI runtime/cost limits.
• Known flaky paths (async timing, shared state, environment coupling).

Deliverables

• Pytest test strategy and case matrix.
• Fixture and parametrization policy.
• Local-fast and CI-full command profiles.
• Residual risk and follow-up actions.

Workflow

1. Build case strategy with assets/pytest-test-plan-template.md.
2. Apply fixture rules from references/fixture-boundary-rules.md.
3. Validate determinism using references/pytest-determinism-rules.md.
4. Define execution profiles in assets/pytest-command-profile-template.md.
5. Finalize with assets/pytest-fixture-stability-checklist.md.

Quality Standard

• Fixtures are explicit, scoped, and deterministic.
• Case coverage includes happy/edge/failure paths.
• Command profiles are reproducible across local and CI.

Failure Conditions

• Stop when fixture design causes hidden shared state behavior.
• Stop when test determinism cannot be achieved for critical flows.
• Escalate when runtime cost blocks practical feedback loops.

Source: KentoShimizu/sw-agent-skills — distributed by TomeVault.