# Corporate Instructor Guide

Release basis: corporate standard reviewed 2026-09-15. Reconfirm the standard, current vendor versions and `FRESHNESS_REPORT.md` before each cohort.

## Delivery pattern

Start each session with due retrieval practice, state one observable objective, explain through an analogy and the real technical model, demonstrate once, supervise safe practice, require verification, then close with reflection and the next workplace connection. Do not lecture through the lab time.

Use the six career chapters as the mandatory route. Assign Oracle, other engines, enterprise HA or data engineering only when the learner’s role or placement evidence justifies it. A cohort may move at different speeds; competency prerequisites remain constant.

## Assessment and answer keys

Lesson and quiz answers are formative and may be discussed after an honest attempt. Never treat client-side question answers as secure examination material. Professional decisions are assessed through observable lab evidence, surprise incident injection and oral defense using the 0–4 rubric in `PROFESSIONAL_CAPSTONE.md`.

For each artifact score: correctness, safety/scope, reproducibility, verification, rollback/cleanup, explanation and communication. Record one specific strength, one evidence gap and the next action. A total score cannot compensate for unsafe conduct or a recovery score below competent.

## Cohort interventions

- **Not started:** confirm access/device and pair for the first ten minutes.
- **Repeated misses:** assign the due review deck and one smaller prerequisite practice; do not unlock later material manually.
- **Reading complete but checks weak:** ask the learner to explain aloud and reproduce without the answer.
- **Lab blocked:** identify whether the blocker is access, environment, command, expected result or confidence; record it rather than marking complete.
- **Moving quickly:** use placement evidence, harder failure injection and peer explanation—not busywork.
- **Inactive seven days:** contact through the approved organizational channel; never expose learner details in public messages.

Export cohort CSV only to approved storage, minimize personal data, restrict access and follow the organization’s retention policy.

## Accommodations

Offer keyboard-only operation, screen-reader-compatible text, captions/transcripts, reduced motion, high contrast/dark theme, extra time, chunked instructions, read-aloud support, alternative typed/spoken evidence and paired setup assistance. Change the access method, not the safety or professional competency standard. Ask the learner what works; do not require medical disclosure.

## Lab facilitation

Before a destructive or external lab, confirm scope, disposable/approved environment, backup or snapshot, stop conditions, expected result, rollback and cleanup owner. During incidents, require observations and hypotheses before commands. Afterward, verify customer-facing health and data integrity, then run a blameless review.

## Release operations

Before each cohort run `npm run audit:freshness`, `npm run test:flags`, `npm run test:tutor` and `npm run test:ui`; inspect `FRESHNESS_REPORT.md`; review vendor-version assumptions; test access, voice fallback, offline mode and one real-box route; record curriculum commit and cohort start date. Escalate broken safety-critical material immediately.

Do not claim that a learner is production-experienced, certified by a vendor, guaranteed employment, or WCAG-certified. Say exactly which evidence level the learner earned: Course Complete, Lab Verified, or Professional Capstone Passed.