All skills
targeted-validation-selection
FoundationJIT onlySelect exact repository-defined validation commands and coverage after a proof strategy exists; use when entrypoint choice or coverage is unresolved.
- Group
- Quality & testing
- Supported roles
- MainAnalysisTaskReview
- Source
- SKILL.md
Use when
Trigger signals that make this skill the right owner
- an accepted proof strategy needs exact repository-defined commands
- repository entrypoints or existing tests have competing coverage claims
Do not use when
Anti-triggers — as binding as the triggers above
- quality-test-gate has not defined proof strategy and observable acceptance
- exact commands and their acceptance and risk coverage are already established
Required inputs
What must be supplied before this skill can decide anything
- accepted proof strategy and observable acceptance
- changed paths and material risk surfaces
- repository guidance command definitions and existing tests
- command targets mutation surfaces hooks/subprocesses credentials external effects authority recovery cleanup and retained-output constraints
- available command results and freshness input/hash/time facts
Output contract
What this skill owes the next role
- Repository-entrypoint inspection evidence covering test/build/schema/lint/static/generator entrypoints and existing tests.
- Record exact smallest-sufficient commands.
- Map observable-acceptance and risk-surface coverage per command.
- Record the expected signal.
- Record command target, working directory, mutation/external-effect classification, credentials/authority, stop condition, recovery, cleanup, and retained-output boundary before execution.
- Record the actual result when run.
- Record freshness input/hash/time facts.
- Record the unavailable-entry fallback.
- State unverified scope, proof limits, and residual risk.
Escalation signals
When this skill must hand the decision back
- no repository-defined entrypoint covers an accepted surface
- an unavailable entrypoint has no evidence-backed repository fallback
- a selected command has unresolved target mutation external-effect credential authority recovery or cleanup boundaries
Loaded by
This skill never owns a task. These professionals may pull it in as Layer 3.
Targeted references
Each reference carries its own load condition. None of them enter context by default.
| Reference | Load when | Do not load when | Required by |
|---|---|---|---|
| repository-command-entry-evidence.mdevidence-pattern | repository-defined entrypoints or existing tests have competing coverage claims | exact commands sources coverage signals and gaps are already recorded | AnalysisTaskReview |