SuperAPI AI Directory

sprint-testing-workflow

Testing & quality

View repository

Updated 10-16 06:50

About this skill

Iterative Testing Workflow (Chinese Version)

English version: See the corresponding English skill.

When to Use

  • When you need to advance according to a sprint/iteration rhythm: planning → setup → execution → regression → stabilization → review.
  • When you need iteration gates and DoD, and hand off to type skills, rather than only doing a single testing task.

Execution Flow

  1. Read and follow prompts/sprint-testing-workflow.md(stages, gates, DoD, handoff).
  2. Supplement with key context such as iteration goals, story scope, capacity, and outstanding defects.
  3. After locating the stage, follow reference.md to name the handoff type skill; relative-path links into other skills' internal files are prohibited.
  4. When information is incomplete, first provide a usable iteration test plan, and mark assumptions and gaps.

Core Constraints

  • Manage iteration stages and exit evidence; hand off complete deliverables to type skills.
  • Gates and DoD must be checkable.
  • When scope changes, recalculate priorities and state the impact on gates.
  • Relative-path links to other skill files are prohibited.

Load on Demand

  • Before producing output, you must read and follow prompts/sprint-testing-workflow.md。
  • Step and handoff cross-reference: read reference.md。
  • Deep work by stage: invoke the corresponding type skill; do not expand complete deliverables within this skill.
  • Template:output-templates/。

Pre-delivery Self-check

  • Has followed the main prompt's output structure
  • Includes stage positioning, gate dashboard, exit criteria cross-reference, and next-hop skill
  • High-risk items have clear priorities and Owners
  • No details not provided by the user have been fabricated
  • Assumptions, trade-offs, and outstanding items are marked

Common Pitfalls

  • Do not write it as a day-by-day log without gates.
  • Do not dump the full body of test cases during the planning stage.
  • Do not claim regression has passed when development has not converged.
  • Do not use "tested a lot" as a substitute for exit evidence.