t1k:scenario
Generate exhaustive edge cases across 12 dimensions before implementation or testing. Use for 'edge cases for X', 'test scenarios', 'what could go wrong', 'boundary testing'.
Files2
View on GitHubSKILL.md.claude/skills/t1k-scenario/SKILL.md
GitHubTheOneKit Scenario — Edge Case Generation
Generate comprehensive edge cases before implementing or testing a feature. Covers 12 orthogonal risk dimensions.
Usage
/t1k:scenario "user registration flow" # All 12 dimensions
/t1k:scenario --file src/auth/login.ts # Analyze a specific file
/t1k:scenario "payment processing" --dimensions 1,2,3,9 # Specific dimensions only
/t1k:scenario "order API" --for-tests # Output as test-plan format
12 Dimensions
| # | Dimension | Focus |
|---|---|---|
| 1 | Auth / Permissions | Who can access, what happens when they can't |
| 2 | Concurrency / Race Conditions | Parallel requests, shared state, locks |
| 3 | Network Failure / Timeout | Partial failures, retries, idempotency |
| 4 | Invalid Input / Boundary Values | Null, empty, overflow, malformed data |
| 5 | i18n / l10n | Unicode, RTL, locale-specific formats |
| 6 | Scale / Volume | Large datasets, high throughput, memory |
| 7 | State Transitions | Invalid state sequences, partial updates |
| 8 | Backward Compatibility | Old clients, schema changes, deprecated paths |
| 9 | Error Cascades | One failure causing downstream failures |
| 10 | Config / Env Variation | Missing env vars, wrong config, multiple envs |
| 11 | Data Integrity | Partial writes, constraint violations, corruption |
| 12 | Observability / Logging | Missing logs, PII leakage, trace gaps |
Detailed scenario lists per dimension: references/dimension-details.md
Output Format (default)
Grouped checklist per dimension with: short description, expected behavior, test approach (unit/integration/E2E).
--for-tests Flag
Output as test-plan structured for the tester agent (tabular format with Test ID, Scenario, Input, Expected, Type).
Pass output directly to /t1k:test as test plan input.
Gotchas
- Skip irrelevant dimensions: A pure CLI tool may not need i18n or auth scenarios — use
--dimensionsto focus - Test approach specificity: Always suggest a concrete test type, not just "test it"
- Cascades are underspecified: For dimension 9, always trace the full failure chain, not just the first hop
- --for-tests output: Verify test IDs are unique across all suites before passing to tester agent
Auto-Activation Keywords
Triggers on: scenario, edge cases, edge case, test scenarios, boundary testing, what could go wrong, risk analysis, corner cases