Stop 9 of 9 · Weekly
Hazem's quarter and the capstone
Three scheduled routines carry the quarter, Cowork clears the documentation backlog, and one client goes end to end against a rubric that scores out of a hundred.
The quarter that runs without you
Eight stops of parts. Here they become a quarter that keeps moving in a bad week.
| When | What it does unattended |
|---|---|
| Monday 06:00 | Reads all twelve tenants, writes one digest: licence drift, stale accounts, compliance-state changes, ranked by risk |
| Friday 06:00 | Diffs this week scan against last week, groups new findings by cause, writes the delta file |
| First working day, monthly | Checks each client roadmap against the asset export and lists what moved, slipped or arrived |
Every one of them reads, ranks and writes a file. None of them changes anything in any tenant. That is not caution for its own sake: a wrong write on a schedule repeats every week until someone notices.
Every Monday at 06:00, using the read-only connectors, produce one digest across the twelve tenants in my tenant list file. Per tenant: licences assigned against licences owned, enabled accounts with no sign-in in 90 days, devices out of compliance, and any change in conditional access or backup job status since last week. Rank the whole digest by risk across all clients, not per client. Flag any tenant where the data was incomplete rather than filling the gap. Change nothing and contact no one.
Cowork takes the documentation backlog: the runbooks from stop 7 that nobody has rewritten in the house style. It works through them while Hazem does the client-facing half of the week, and every inferred line still waits for a person.
The capstone
Pick one real client. Run this course end to end on them this quarter and keep the evidence. Five artefacts: the roadmap and priced budget, the frame mapping and gap list, the remediation plan and client note, the runbook with its tabletop after-action, and the redaction log recording every file you attached and what you stripped from it.
[attach the five capstone artefacts and the redaction log] Grade this against the rubric below, out of 100, scoring each criterion separately and quoting the evidence you used. For every point not awarded, say exactly what is missing and what would earn it. Be strict about dates that came from you rather than from a source, about gaps with no owner, and about any file in the log that was attached without redaction.
The rubric
| Criterion | Points | Full marks means |
|---|---|---|
| Redaction integrity | 25 | No credential, secret, raw scan output, patient or case data in any transcript, and a log line for every file attached |
| Roadmap defensibility | 20 | Every line has a dated trigger, every price came from a quote, undated assets sit in their own section |
| Gap list is actionable | 20 | Every gap cites a control from the attached document and carries an owner, a cost band and a date |
| Remediation by cause | 15 | Findings grouped to causes with owner and window, and no unsourced claim in the client note |
| Runbook survives rehearsal | 20 | No blanks left, twelve injections run, every unanswered question now a step with a named role |
Eighty passes, and redaction integrity is pass or fail on its own. Under eighty the gap is nearly always the same two: dates nobody verified, and gaps with no owner.