Quick Reference: Domain 6 — Stakeholder Communication & Lifecycle Management

Domain 6 distractors come from four recurring postures. Name them and roughly nine items become mechanical. No Anthropic document describes this taxonomy and no item will use the words.

The keyed answer is almost always the fifth option: measure it, frame it, recommend, and leave the decision where the accountability sits. When misses cluster by posture rather than by objective, drill the postures directly.

The request is not the requirement. The stated solution is evidence about the problem, never a substitute. Building first produces a demo on cooperative data that anchors every later conversation; pricing first quotes work nobody has scoped.

Discovery is judged by its outputs, not by how many people were consulted:

• Anthropic's criteria standard (define-success): specific, measurable, achievable, relevant, with targets from benchmarks, prior experiments or expert knowledge. Its worked contrast replaces "the model should classify sentiments well" with a named held-out set of stated size, a quality floor, a safety rate, an inconvenient-versus-serious error split, and a response-time bound. Most use cases need several criteria, not one headline number.
• Feasibility triage is discovery, not a later surprise: a decidable task with a verifiable output; data that exists, is good enough and may lawfully be used; a cost of error tolerable at an achievable accuracy. Saying no honestly, with the evidence, is a valid discovery outcome and can be the right answer to a scenario, and Anthropic's agent guidance agrees from the other side — find the simplest solution possible, adding complexity only when it demonstrably improves outcomes.
• Distractors: the clickable demo, the fixed price, the ranked wish list from a workshop, and deferring data quality or lawful basis to design.

This objective is almost entirely professional craft rather than documented product behaviour. Its authority is the guide's description of the role: collaborating with cross-functional stakeholders and communicating architectural decisions effectively, in rooms containing security, legal and executive participants.

The seven-part communication (ours): the decision · the options genuinely considered, including the one the audience expected · the criteria they were judged on · the evidence, which for a Claude system means measurements rather than assertions · what each option costs and exposes · the recommendation, because declining to recommend is abdication rather than neutrality · what would change the recommendation. The last part separates an architect from an advocate, and it is the content a decision record carries in 6.4.

• Translation changes the lead, never the facts. Executives lead on consequence and risk; engineers on mechanism and constraint; legal on obligation and residual risk; the sponsor on cost, date and what they are accepting. Burying an executive in mechanism and handing a specialist generalities are symmetrical distractors. Simplifying past the point of truth is worse than either, because whoever trusted it finds out later.
• Probabilistic honesty is an obligation. State the error rate, the population it was measured on, and the handling path for the residual errors. A demonstration is a communication risk rather than a tool: a sample from the favourable tail, implying a determinism the system does not have.
• Uncertainty travels as an interval. Anthropic's evaluation research supplies the register: report the standard error and a confidence interval with each score, and give pairwise differences their own intervals, because a comparison without them may be presenting noise as a finding. A single confident number offered as the accuracy answer is a distractor even when plausible.

The most frequently keyed shape in the domain: an accountable stakeholder wants what the evidence contradicts. Measure, frame, recommend, and leave the decision where the accountability sits — the quality difference as a number, the downstream cost in the currency the owner controls, the exposure, the saving, a clear recommendation, then their decision.

Read the stem for the consequence, not the volume of the stakeholder. Escalation is the last move, and only where the owner's authority does not cover the consequence.

Service levels are engineered, then promised. Never the reverse. The commitment comes from a measured envelope on the real pipeline with the real corpus under realistic concurrency, with headroom for load the measurement never saw. A number that travelled the other way is a liability with a date on it.

Six elements. A stem offering a bare figure is offering a distractor:

Anthropic's success-criteria example reaches the same register from the quality side — a named held-out set of stated size, a quality floor, a safety rate, a response-time bound. The six-element framing is ours, drawn from established service-level practice.

Negotiating the impossible ask is the most commonly keyed situation here. Show the decomposition so the figure stops being arbitrary, state the envelope the pipeline does hold, then attack the gap where it can honestly be attacked: perceived responsiveness. Anthropic's latency guidance separates baseline latency from time to first token and recommends streaming because it improves how responsive an application feels. Progressive disclosure, partial results and an asynchronous path with notification are the same lever on other surfaces.

• None of those levers moves total completion time. Where the consumer is a parser or a batch job rather than a waiting person, they are a distractor.
• Two moves are always wrong. Accepting a number the pipeline cannot hold is capitulation. Hitting it by deleting the retrieval or verification steps the answer depended on is worse: it trades a measured problem for an unmeasured one, and the promise met is no longer the promise made.
• Quality gets identical treatment. "It will be very accurate" fails as a bare duration fails. Where the shortfall concentrates in a segment, say so, commit on the population genuinely measured, and give that segment its own improvement plan, evidence and owner.
• The trap that looks like rigour: quietly narrowing the population until the headline number becomes true. The excluded cases stay in production with no threshold, no review path and nobody accountable. A keyed option may narrow a population, but only where the exclusion is visible and nobody is left uncovered by accident.

Dated. Re-verify against Anthropic's service-tier documentation before writing any availability clause; the tiers, the target and the commercial terms have all moved before. The point survives them: a promise written above the envelope beneath it is written against somebody else's best effort.

• The commercial caveat is easy to miss: Priority Tier depends on an existing capacity commitment, and as of mid-2026 capacity commitments are documented as no longer available for purchase to organisations that do not already hold one. Do not assume the tier is available to a new client.
• Traffic beyond a committed capacity falls back to standard tier automatically — the promise degrades exactly when volume is highest. It is also a worked example of a designed breach path.
• In a negotiation: check the tier the system actually runs on, and mark anywhere the promise exceeds the envelope beneath it.

A feedback loop is instrumentation, not sentiment. It earns the name by producing evidence and disposition.

• Evidence: the channel converts a report into something workable — the trace, the case identifier, the reproducible input. "It gets things wrong sometimes" cannot be actioned; a trace can.
• Disposition: triage has a visible outcome, or the channel trains everyone to stop reporting and silence gets mistaken for quality.
• Permanence: confirmed failures route into Domain 4's regression set. Anecdotes routed straight into prompt edits is how a system rots, each edit fixing the reported case and moving one nobody mentioned.
• Relevance beats frequency. Reporting is led by the criteria discovery agreed, with engineering measurements underneath as support. A sponsor who agreed to a business outcome and receives infrastructure telemetry has stopped being communicated with, and the fix is relevance, not a shorter interval or a longer deck.
• Scope change is the same discipline pointed at the plan: weigh each request against the agreed criteria and baseline, put cost, date and risk to the sponsor before anyone commits, and reset the plan under their signature. Silent absorption and blanket refusal both fail; absorbing scope by cutting testing that nobody agreed to cut is the worst option available, and the exam plants it.

Documentation transfers judgement, not history. The next person needs what was decided, on what evidence, what consequences were accepted, and how to tell whether they have broken something. The test for every page is whether it changes what a competent stranger would do.

The artefact set is established practice in our voice. Decision records and runbooks are long-standing industry artefacts, not Anthropic products.

• Two questions test a pack (ours): can a competent engineer who was not there change the system safely, and will they know when they have broken something. Anything answering neither is there for the author. Page count answers neither; a recorded walkthrough answers the first only while the system still resembles the recording.
• The baseline is the keystone, the only artefact answering the second question. Anthropic's evaluation guidance gives its design standard (task-specific, edge cases included, automated where possible, volume prioritised over hand-grading); its research gives the comparability standard (report the interval with each score, so a re-run reads as better, worse or indistinguishable rather than merely different). Domain 4 designs the suite; packaging it as transferable is here.
• Discoverability is part of correctness. Drifted documentation is worse than none because it is believed. Keep it beside the code, version it with the code, change it in the review that changes the behaviour. A thorough wiki in a tool the team never opens is a private document.
• Implementation guidance is more than a diagram: the sequence, the risky steps and therefore the order that is not negotiable, the gates, a definition of done for each. Guidance that names the components and stops has handed over the part that was never in doubt.

Phases are gated by what they produce, not by dates. Output gating is our framing; nothing turns on reciting phase names, and no item may reward project-management vocabulary.

• Ownership transfer is capability transfer. Four abilities decide whether handover happened: run the evaluations and get a comparable answer, read a trace well enough to localise a failure, roll a change back, and know who to reach when none of that is enough. A team missing any of them will operate the system and refuse to change it, which looks like stability and is paralysis. Domain 7 does the enabling; Domain 6 decides what must be true and verifies it.
• Iteration is evidence-gated. Prompt edits shipped against last week's complaints with no regression suite is drift with optimism attached.
• Moving backwards is professional. Production evidence contradicting a discovery assumption sends the work back to discovery. Treating the phase order as a one-way ratchet is the distractor.
• Quiet is not evidence. A system nobody has complained about, whose baseline has not been re-run and whose corpus has moved underneath it, is unevidenced rather than healthy.

Dated. Re-verify the states, the notice period and the checklist against Anthropic's model-deprecation and migration documentation before planning against them. The architectural reading does not move: a deprecation is a scheduled re-validation, not an incident.

• Notice: Anthropic commits to notifying customers with active deployments at least 60 days before retiring a publicly released model. Treat that as the planning horizon, not the working window.
• The documented moves: test with the replacement well before the retirement date, and audit usage by API key and model to find where the old version is still called. A pinned version in a forgotten job is the usual surprise.
• Re-benchmarking cost and latency is on every migration path: a documented checklist item in its own right, alongside the behaviour changes each path calls out for explicit re-testing. It is neither the last step nor a courtesy, and it is what makes the next comparison mean anything.
• The same treatment applies to events with no notice period: corpus migrations, connector and dependency changes, regulatory change, and the redesign of the business process the system serves. Each forces re-validation, and each is cheaper planned for in design.
• Distractor: handing a deprecation to the commercial or vendor-relationship owner. The date and the replacement are commercial facts; re-validating against the baseline is architectural and cannot be passed on.

Half the wrong answers here are right answers to a neighbouring domain's question.

Item formats and our conventions. The guide names multiple-choice and multiple-response, and states that every multiple-response item says how many responses to select; it states no partial-credit policy. We additionally drill scenario-matching, several short scenarios classified against one shared reusable option set, because it tests the same judgement under tighter conditions. That format, all-or-nothing scoring, and the positional scoring we apply to matching drills are our practice conventions rather than stated exam rules.