Coverage Check (test case)

Instructions

Audit the test case $ARGUMENTS against the journey test and the code that are supposed to realize it, and report the result.

This skill is a front door to the read-only uc-coverage-layered sub-agent of aiup-core, the same agent /use-case-coverage uses. The audit checklist lives in agents/uc-coverage-layered.md at the plugin root (locate it with a glob for **/agents/uc-coverage-layered.md) and is deliberately not repeated here. How a report is presented, compared with an earlier run, and run on a host without sub-agents is owned by /use-case-coverage: locate it with a glob for **/*use-case-coverage/SKILL.md (the folder may carry a host prefix such as tessl__use-case-coverage) and apply its sections "Delegation", "Repeated Runs", and "Hosts without sub-agents" as written. Your job is argument parsing, delegation, faithful presentation, and offering the next step.

The report is the deliverable. You do not fix what it finds.

Arguments

Everything is parsed out of $ARGUMENTS; the tokens may appear in any order.

A mode is only narrowed by a standalone qualifier token, never by a word in a sentence. State the resolved arguments in one line before delegating (Auditing TC-001, mode both.).

Workflow

1. Parse $ARGUMENTS into the id, the mode, and the work-in-progress flag.

2. Confirm the test case exists — docs/test_cases/TC-XXX-*.md. Stop, list the near matches, and ask if it does not. Never audit against a journey you inferred from the tests.

3. State the resolved arguments in one line.

4. Delegate the audit to uc-coverage-layered (it may appear as aiup-core:uc-coverage-layered), passing exactly the id, the mode, and — when it applies — work in progress, nothing else:

   TC-001 both
   TC-001 tests work in progress

5. Present the returned report unchanged, and on a rerun add the comparison — the sections of /use-case-coverage named in Instructions.

6. Offer the gap-closing commands — see After the Report. Then stop.

With no id: if the conversation has just been working on a specific TC-*, propose that one and ask. Otherwise list docs/test_cases/ with each test case's **Status:** line and ask which to audit. Audit one journey per run.

After the Report

• A test gap — a Flow row or Validation item without a test step — closes with /test-case-implement TC-XXX.
• An implementation gap sits in a use case the journey chains: name the use case of that Flow row and offer /use-case-implement UC-YYY, narrowed to a layer when the report names one.
• Missing Markers rows of the E2E layer close with a rerun of the E2E test skill named in the Layers table; docs/architecture/stack.md itself missing closes with /stack.
• Drift is not a gap with a known fix. For each drift item, ask the user which side is right — the test case (/test-case TC-XXX updates the journey) or the application (/use-case-implement UC-YYY) — and offer only the command that matches the answer. Never pick a side yourself.
• Offer the commands; run one only if the user says yes. Pass an Automated suggestion on as a suggestion: it says the journey is automated, not that it passes, and the agent cannot run it — never claim it was run. Do not edit the test case's **Status:** line.

DO NOT

• Do not write or edit code, tests, or specifications — including the **Status:** line.
• Do not restate or paraphrase the agent's audit checklist in this file; agents/uc-coverage-layered.md owns it.
• Do not soften, upgrade, or drop a verdict, and do not present a summary in place of the matrix.
• Do not tell the agent what you expect it to find.
• Do not claim that a build or a test suite was run.