Find the step that broke a Ghost Inspector test, and why

**Objective:** Find the cause of the most recent failure of the Ghost Inspector test named `{test-name}`, looking back `{lookback-days}` days (default 7), and report the failing step with supporting evidence.

**Step 1: Identify the test**

Call `ghost-inspector-get-test` with `name.contains` set to `{test-name}`. If more than one test matches, list the matches with their suite and id and stop, asking which one is meant. If nothing matches, say so and stop.

**Step 2: Find the failing run**

Call `ghost-inspector-get-test-result` filtered by the `test-id` from step 1, `passing` false, and `date-created.from` set to the start of the lookback window. Take the most recent result. Record its `id`, that is the test result id every later step needs. If there are no failing runs in the window, say the test has not failed in that period and stop.

**Step 3: Locate the failing step**

Call `ghost-inspector-get-test-step` with `test-result-id` set to the id from step 2 and `passing` false. Then call it again without the `passing` filter to get the full sequence, so you can say what the test had done up to that point.

* the `command`, `target` and `value` say what the step tried to do
* the `sequence` says how far through the test it got
* a failing assert or click usually means the page did not reach the expected state, not that the assertion itself is wrong

**Step 4: Check the browser console**

Call `ghost-inspector-get-console-log` with `test-result-id` set to the id from step 2 and `error` true. Match any errors against the time and url of the failing step. If there are no errors, call it again without the `error` filter and look for warnings around the same url.

**Step 5: Check whether anyone has already looked**

Call `ghost-inspector-get-test-result-comment` with `test-result-id` set to the id from step 2. If someone has already commented, quote it rather than repeating their analysis.

**Step 6: Report**

* Test, suite, browser and region
* When it last failed, and the run url it started from
* The failing step: sequence number, command, target, value
* What the test had successfully done before that point, in one or two lines
* Console errors that look related, quoted
* Your best explanation, and whether this looks like an application defect, a timing problem or a stale test

**Rules and guidelines:**

* This is a read only investigation. Do NOT call `ghost-inspector-execute-test`, `ghost-inspector-execute-suite`, `ghost-inspector-update-test`, `ghost-inspector-duplicate-test`, `ghost-inspector-add-suite-result-comment` or any other tool that changes or runs anything.
* Do not guess at a cause the evidence does not support. If the step data and the console are both clean, say the failure is unexplained by the captured evidence and recommend a rerun.

**Validation:**

* If `{test-name}` matches nothing, report that and list a few test names from `ghost-inspector-get-test` so the user can correct it.
* If the step list comes back empty, say the run recorded no steps and report only what the result row itself carries.