scenario-test-lab

star 0

Build, extend, and execute reusable dummy-data scenario tests under tests/ for duplicate-serial and backup-collision workflows. Use when continuing scenario-based validation (for example S6/S7 style additions), delegating scenario work to multiple agents, and enforcing strict legacy safety (never modify legacy code/files).

Rukkha1024 By Rukkha1024 schedule Updated 3/8/2026

name: scenario-test-lab description: Build, extend, and execute reusable dummy-data scenario tests under tests/ for duplicate-serial and backup-collision workflows. Use when continuing scenario-based validation (for example S6/S7 style additions), delegating scenario work to multiple agents, and enforcing strict legacy safety (never modify legacy code/files).

Scenario Test Lab

Enforce Safety First

  • Treat every legacy code/file as read-only.
  • Never modify production files or existing legacy files.
  • Limit write scope to new or explicitly approved test artifacts and records under tests/.
  • If file ownership is unclear, pause and ask before editing.
  • Run conda run -n module ... serially only; never run conda commands in parallel.

Follow This Workflow

  1. Inspect current scenario baseline.
  2. Define goal-oriented scenarios with explicit pass/fail checks.
  3. Delegate scenario design and verification to multiple agents in parallel.
  4. Implement new scenarios in test scope only.
  5. Execute scenarios serially with conda run.
  6. Record outcomes in a persistent scenario journal.

Build Scenario Backlog

Use this minimum matrix and keep extending it.

  • subject_info duplicate serial, same wear date
  • subject_info duplicate serial, different wear dates
  • backup duplicate with matching.exhausted_policy=skip
  • backup duplicate with matching.exhausted_policy=error
  • mixed batch with success + skip + error coexistence

When adding more cases, follow S<number>_<short_name> key style and keep old scenarios stable.

Delegate With Agents

Use parallel agents for design, expectation definition, and review.

  • Agent A: draft dummy rows/files and edge-case intent
  • Agent B: draft expected statuses/log tokens/final filenames
  • Agent C: review scenario consistency and overlap gaps

Then consolidate once and run commands serially.

Execute Scenarios

Use repository-local runner first (if present), then fall back to direct command.

Keep Persistent Records

Initialize or append scenario journal entries under tests/.

Use References

  • For safety boundaries and mutation policy, read references/legacy-safety-checklist.md.
  • For scenario catalog patterns and naming, read references/scenario-catalog.md.
Install via CLI
npx skills add https://github.com/Rukkha1024/kakaocli_skill --skill scenario-test-lab
Repository Details
star Stars 0
call_split Forks 0
navigation Branch main
article Path SKILL.md
More from Creator