flowguard-test-mesh

name	flowguard-test-mesh
description	Use when tests, checks, transition cells, payload cases, or evidence are large, slow, layered, stale, skipped, backgrounded, release-only, or require parent/child ownership and freshness proof.

FlowGuard Test Mesh

Purpose

Govern parent/child test hierarchy, validation partitions, results, and freshness.

Entrypoint Scope

This is a standalone FlowGuard satellite skill; test_mesh_maintenance owns evidence structure, not semantics, process optimization, or test execution.

Local Material Routing

Read references/test_mesh_protocol.md for split, ownership, diagnostic accounting, reuse, matrices, and release scope.

Entrypoint Acceptance Map

Review a model-derived parent/child validation mesh; block stale, skipped, incomplete, or unowned evidence; hand semantics and lifecycle/risk to typed owners.

Use When

• Use for large/slow/background child test scripts, stale/reused evidence, release gates, artifact-payload matrices, or diagnostic boundaries.

Do Not Use When

• Do not split code/models, choose DPF process shape, group root causes, decide semantics, or execute tests; return small tests to model-first-function-flow.

Required Workflow

1. Define the parent gate and derive child suites/scripts from a FlowGuard validation-structure model.
2. Declare an independent inventory revision and every required surface, obligation, member, cell, case, and shard; map each id to one owner.
3. Attach status, freshness, artifacts, reuse tickets, skips/timeouts, terminal identity, fingerprint, covered ids, and versions. Diagnostics add diagnostic_boundary, planned/executed/failed/not-run counts, not-run reason, campaign id, and stable Finding Ledger ids. Provider checks preserve session/consumer and receipt identity.
4. Review routine/release scope and return child evidence plus typed handoffs.

Hard Gates

• Model-purpose gate: before build/change, freeze this instance's task-specific failure(s) and boundary; then bind candidate plus native good/bad-per-failure/oracle/current evidence. Reusable types are not fixed-purpose; no mode/fallback; SkillGuard only supervises FlowGuard-declared checks.
• Use the real FlowGuard check engine and AGENTS.md managed record; never create a fake mini-framework.
• PID/log/running/progress proves liveness only; reuse requires current TestResultReuseTicket and ProofArtifactRef.
• One receipt may fan out but cannot be copied or counted as several executions.
• Require planned = executed + not_run, failed <= executed, no not-run under declared_complete, visible reasons elsewhere, and stable finding ids for failures.
• Locally green subsets cannot prove an independently declared complete inventory.

Output Requirements

• Return evidence, failures, blockers, skipped_checks, residual_risk, claim_boundary, typed_next_actions, a validation mesh diagram, and child freshness.

SkillGuard Maintenance

• Edit contract source and regenerate; SkillGuard cannot turn liveness into pass.