# The unified teaching workflow

Updated 24 September 2026. This records Chad's intended workflow and the remaining work. It does not change the workbook standard, course sequence or existing evidence.

## What the system is for

Choose a course and sprint in Evidence Map → open its assignment package → copy its course-specific brief to the existing ChatGPT workbook maker → return the workbook, coach context, checks, rubric and registration files together → teach through GitHub → collect Forms and iPad evidence → review the student's criteria, skills and outcomes in Evidence Map → reuse the teacher's comment and confirmed judgment in PowerSchool.

Truck Ad remains the working instructional benchmark. Learning Studio Workbook Standard v2 records its reusable structure together with the later Drop Day/Fresh Impact refinements. Do not replace strong teaching with a generic template or rename the system again.

The package, not an individual app screen, is the unit that must be complete. Google Sheet owns the roster and Forms intake. Evidence Map is the main dashboard. The iPad is its field-capture companion. GitHub holds published teaching materials. ChatGPT builds teaching packages from the same specifications. The coaching Gem supports students within each package's assessed-thinking boundaries. PowerSchool receives reviewed reporting information.

## What the audit found

- The connected backend and live roster were functioning, but the older app's assignment catalogue and workbook-brief controls were not carried into the connected dashboard. That created a real workflow regression.
- The current plan has 9 MM12, 9 COM11 and 16 IBDS sprint assessment bundles. These remain planning records; not all student materials, labs, teacher briefings or task rubrics have been authored.
- The older local catalogue has 111 assessment specifications from an earlier plan. It remains preserved in the local planning tools. Those entries must not silently overwrite the later dated course plan.
- Truck Ad's existing Google project record points to the `technology-learning-system-v3` branch of `ctwadden/classroom-tools`. Its active student/teacher Forms and six Form checkpoint definitions exist. They should be reused.
- The original Truck checkpoints include Smart Filter mask repair, blend-mode reasoning, typography/hierarchy revision, audience/message defence, the final advertisement, and critique/revision. The source evidence map also describes layered-file and export checks. The Netlify rubric summarizes the task into three broader criteria. There is not yet a reviewed, explicit checkpoint-to-criterion map joining these versions.
- The original rubric includes help-dependent lower-band wording. That must be reconciled with the agreed rule that achievement and support are separate. Do not silently rewrite historical captures or approve a rubric on the teacher's behalf.
- The MM12 plan lists 21 skills for Explorer/Truck Ad. The current shared Truck rubric explicitly tags only two skills. The older crosswalk also records outcome-only MM12 joins. Skill coverage therefore needs reconciliation before it can be presented as a complete skill profile.
- The public Drop Day page has working workbook/coach/assessment links, but its step 1 still displays `__DROP_ASSET_A002__`. Record this as a workbook repair; it is not evidence that the whole standard needs replacement. The current GitHub Truck manifest also labels the package as a release candidate awaiting school preflight.

Sources inspected: the Google **Projects** and **Form Registry** tabs; the published Drop Day page; Truck Ad's [project manifest](https://github.com/ctwadden/classroom-tools/blob/technology-learning-system-v3/system/projects/photoshop/truck-ad/project-manifest.yml), [Forms specification](https://github.com/ctwadden/classroom-tools/blob/technology-learning-system-v3/system/projects/photoshop/truck-ad/assessment/forms-spec.json), [original rubrics](https://github.com/ctwadden/classroom-tools/blob/technology-learning-system-v3/system/audits/photoshop/truck-ad/TRUCK_AD_EVIDENCE_RUBRICS.md), and [evidence map](https://github.com/ctwadden/classroom-tools/blob/technology-learning-system-v3/system/audits/photoshop/truck-ad/EVIDENCE_MAP.md). Local implementation: connected.js, connected-dashboard-core.js, connected-catalogue.json and the existing course map. GitHub branch inspected at commit `3b7fc04d3141fb43e3aaa165e5831a14fd36b796`.

## What was restored in this update

The existing visual dashboard now has **Assignments & workbooks**. Every current sprint bundle exposes its lesson flow, outcome references, planned skills, O/C/P opportunities and a copy/download workbook-maker prompt. Explicit project registrations join known resources, rubric versions and existing classroom evidence by course and project ID. Missing links remain visible. Source checkpoint identities are included in the prompt.

Rubric cards now show their existing collection guidance and linked skills. Assignment evidence shows the same server records as student review, including knowledge and reflection streams. Skill counts show explicit tags on eligible records; they are not mastery scores. The dashboard now copies individual evidence notes and confirmed reporting comments without rewriting them. These controls do not post to PowerSchool or change grades.

`teaching/assignment-links.json` is the small explicit relationship registry used to build the existing connected catalogue. It introduces no separate student database. The build regenerates the catalogue from its source files. It is not yet a teacher-facing import/publish workflow for returned workbook packages.

## Remaining work, in order

1. **Finish Truck Ad as the reference package.** Register the current published workbook and its coach context; reconcile the original checkpoint, criterion, outcome and skill mappings; retain one reviewed rubric release. Resolve the difference between the original manifest's outcomes and the wider Explorer/Truck sprint outcomes without treating every planned opportunity as assessed evidence. Repair and rehearse the actual files.
2. **Finish the package return workflow.** A returned workbook package must update its existing assignment record with versioned files, published links, Forms specification, rubric and mapping validation. Build the review/import/publish step so a new package does not need manual reconstruction across apps. Source editing currently supplies these registrations; this final authoring workflow is not implemented.
3. **Finish student criterion/skill review and the device rehearsal.** Keep each capture attached to the exact rubric version, criterion, taught skill and evidence method. Show coverage, evidence and teacher judgments separately. Test actual iPad Safari capture, airplane mode, reconnect, duplicates and revisions. Confirm one capture and its exact Google mirror receipt. Recording bytes still stay on the device; Gemini requires configuration.
4. **Finish reporting with the school setup.** Copy/paste confirmed comments works; the CSV is a review register. Verify the actual PowerTeacher exported template, student numbers, local standards and accepted scale. Add a teacher-confirmed transfer record so the app can distinguish prepared, transferred and revised information. A copied comment must not silently mark itself submitted.
5. **Produce the remaining packages using that proven path.** Keep the established MM12/COM11 semester and IBDS September–May sequence. Create and QA content in sequence, including the IBDS chapters, source work, labs, papers, IA preparation, mocks, decks and daily teacher notes. Do not count planned rows as finished instruction.

## The test that establishes completion

For one fictional learner and the finished Truck package: open the actual workbook and correct check → route the submission to the intended project/course → record one iPad observation and conversation using the same released rubric → see the notes, criterion and skills on the assignment and student dashboard → obtain the exact-revision Google receipt → make a deliberate outcome judgment → copy the exact reviewed comment → validate the school-supported reporting handoff. Refresh/retry must not duplicate the record. A revised rubric must not alter the interpretation of the older capture.

Only then repeat the package workflow across the courses. A successful build, authentication check or roster sync is a component check, not this completion test.

## PowerSchool limit that affects the design

PowerTeacher Pro's documented CSV import targets **one assignment at a time**. It does not establish a single upload that creates the course's assignments and imports all independent standards judgments and comments. Comments have a documented manual entry/comment-bank path. Use the copy controls now; validate any further automation against the actual local setup. See [score import/export](https://ps.powerschool-docs.com/powerteacher-pro/latest/importing-and-exporting-scores) and [comments](https://ps.powerschool-docs.com/powerteacher-pro/latest/enter-scores-and-comments).


## Existing Photoshop Bootcamp teaching files

Open [MM12 Bootcamp](https://outcome-evidence-map.netlify.app/connected.html?course=MM12&view=assignments&sprint=mm12-26-s01) or [COM11 Bootcamp](https://outcome-evidence-map.netlify.app/connected.html?course=COM11&view=assignments&sprint=com11-26-s01). The original guides are listed under **Your teaching resources**:

- Classes 1–3: Below the Surface — existing student HTML and saved PDF; embedded shark source, illustrated steps and motion reference.
- Classes 4–6: Drop Day — existing published GitHub workbook, coach and assessment launcher.
- Classes 7–9: On the Cover — existing student HTML and saved PDF; embedded portrait, supplied copy and construction references.
- Optional prerequisite/support: Photoshop Foundations — Decode the Impossible, with gallery, simulator, phone portal and out-of-bounds effect.

The first and third guide were recovered from existing files, not rewritten. Their local responses do not yet feed the shared evidence system. Exact coach/Form/shared-rubric registrations remain unfinished; class 10 is still a planned independent skills check.