Find the missing responsibility
Silent video.

[0–6]
Describe the blocked task • Start with a release or incident you cannot operate. • Existing server + release workflow • W17 · new worker • Deploy step + retry test • Who owns this task, and does the agreement already cover it? • Invented worker example · describe the actual blocked operation
In the invented application example, an existing server and release workflow remain visible while new worker W17 connects to a required deployment step and retry check. Ask who owns that concrete task and whether the current agreement covers it. Name the queue, release sequence and expected failure behaviour. Hosting and release-process work can overlap; service names alone do not define acceptance, operating coverage or ownership.

[6–12]
Read the existing scope • Hosting and application agreements may already cover it. • Requested task • Existing agreement • Included work • Defined addition • Existing inclusion → complete or use it • Uncovered work → define only the addition • Check proposal, conditions, environments and allowances together
The complete reviewed figure sends a requested task through the existing agreement before considering included work or a defined addition. Existing inclusions should be used or completed; missing documentation can be a handover correction rather than a new purchase. Only genuinely uncovered work needs a defined addition. Read current service scope, proposal, conditions, allowances and covered environments together. Existing hosting, cloud and application release and recovery deliverables remain intact; no quoted price or plan is changed.

[12–18]
Name the remaining gap • Define the extra task and its acceptance evidence. • Keep existing server + workflow scope • Review uncovered W17 work • W17 · proposed extension • Queue + sequence + retry evidence + owner • Incompatible records may need application work; a tool cannot decide it
The existing server and workflow remain fixed. Only the proposed W17 extension moves into the review lane, keeping its identity below the review label. Confirm the actual uncovered task and define its queue, deployment sequence, retry evidence, acceptance and ongoing owner. If the worker cannot safely read records from the previous version, application compatibility work may be required; a bigger server or different pipeline cannot decide this alone. No additional work is commissioned in the film.

[18–24]
Agree who responds • Alerts need recipients, cover and escalation. • Monitoring signal • Alert recipient • Agree people + hours + cover + decision authority • Response coverage • Escalation • An alert does not establish 24/7 response
The reviewed monitoring-signal-to-recipient track appears first. A native boundary then calls for explicit people, hours, cover and decision authority before the separate response-coverage and escalation track appears. An alert does not establish round-the-clock incident response, automatic rollback or an availability commitment. Compare actual support terms, maintenance and escalation. Backup schedules, restore-rehearsal evidence and recovery-time commitments are different facts.

[24–30]
Test the handover • A maintainer must locate evidence and act on failure. • Versioned release evidence • Named maintainer + working access • R17 • Failed check → agreed recovery / escalation route • Verify account ownership, transfer, instructions and actual rehearsal
Illustrative release evidence R17 moves from its versioned source to a named maintainer with working access, keeping the reference unchanged. The next action is the agreed route for a failed check, to be rehearsed with actual acceptance evidence. Verify the deployed version, access and recovery instructions, account ownership, billing, credentials, backups, workflow files, contacts and transfer responsibilities. Test environment separation too. Compare infrastructure usage, licences and third-party costs separately from setup, migration, management and application changes; scope additional work without reclassifying existing inclusions.