LLND digital instrument — production pack content draft (digital v2) — DRAFT FOR TIM REVIEW¶

Status: DRAFT — nothing publishes from this document. The prod publish op is a separate, later brief gated on Tim's explicit nod (commissioning ruling 2). This is the Tim + Claude content workstream artefact: the RTOpacks-authored, framework-mapped production item set.

Authored: 2026-07-10, RTOpacks-side Claude (helm session), for Tim's item-by-item copy review. Pairs with: llnd-02-spec §4.1/§4.3/§9.5; ruling-2026-07-10-conduct-commissioning-batch; llnd-source-captures-2026-07-09 §1.

1. Provenance — what this was authored against¶

Framework source: DigComp 2.2 (JRC128415; Vuorikari, Kluzer & Punie, JRC 2022). Authoring reads were on the project-files copy of JRC128415_01.pdf — the same bytes Tim digest-pinned on capture (sha256 9733ab7946f094be78c00f6a839e26be5486a00d6f953d93ddf42a5c135ab359, llnd-source-captures §1). Every item below anchors to a Dimension-3 proficiency-level statement and one or more Dimension-4 KSA examples, cited by the framework's own example numbers.
Cross-check against the in-custody extraction: DONE (2026-07-10, same session). Authoring initially mapped against the verified PDF bytes (repo unreachable at draft time); Tim then connected the repo and the check ran on bytes: sources/frameworks/digcomp-2.2-jrc128415.pdf sha256-verified identical to the pinned reference digest (9733ab79…); all 39 KSA example numbers cited below verified against sources/frameworks/digcomp-2.2-extraction/digcomp-2.2-ksa.csv — every citation sits under the competence it is anchored to (one cross-competence side-citation found on d02 and removed in this revision); the descriptors CSV carries the full 21 × 8 level matrix consistent with the level statements quoted in the anchors.
Substrate grounding (live, 2026-07-10): pack shape and scoring contract read from rtopacks-llnd-dev on bytes — llnd_instrument_pack DDL and the pilot pack pack_235a36e7 (digest 49f2487f…). The production pack below is shape-identical to the pilot: same top-level fields, same item fields, same level_mapping structure (count-of-correct → level, per competence). No new field is introduced — the deployed conduct surface's candidate-safe projection and the deterministic scorer are consumed as-built, untouched.

2. Design decisions taken at the helm (D1–D7) — Tim can reverse any of these before publish¶

D1 — Competence subset: 10 of 21, all five areas. The pilot's eight (1.1, 1.2, 2.1, 2.6, 3.1, 4.1, 4.2, 5.1) plus 1.3 (managing data/files — saving, organising and retrieving work is the single most load-bearing digital act in online VET study) and 2.2 (sharing through digital technologies — submitting work and basic attribution). This is the pre-enrolment-relevant subset the spec's §4.3 mapping note anticipates. Excluded and why: 2.3/2.4/2.5 (citizenship, collaboration, netiquette — training-phase behaviours, not enrolment-suitability gates); 3.2/3.3/3.4 (re-elaboration, licensing, programming — beyond any foundation VET product's pre-enrolment demand); 4.3/4.4 (wellbeing, environment — real but not suitability-screening territory); 5.2/5.3/5.4 (tool selection, creative use, gap identification — partially covered by 5.1 at this grade; candidates for a v3 if a product bar ever demands them).

D2 — Measurement range: levels 1–4 only (foundation + intermediate). The instrument does not claim advanced. DigComp levels 5–6 are defined by "guiding others" and "complex contexts"; levels 7–8 by creating solutions to complex problems. A constrained-response screen cannot evidence any of that — a selected-response answer key asserting level 6 would be indefensible at audit. The pilot's mapping reaching 6 was placeholder machinery-exercise, not precedent. The production scale per competence: 0 correct → 1, 1 → 2, 2 → 3, 3 → 4. Floor semantics, documented honestly: a reported 1 means "not demonstrated above basic-with-guidance" — the screen cannot distinguish level 1 from below-1 (both DigComp 1 and 2 are "basic level"; they differ only in guidance-dependence, which an unproctored screen cannot observe). Ceiling semantics: a reported 4 means "intermediate demonstrated; the screen does not measure higher." Config guidance follows: product digital_bar values should sit at 2–4; a bar above 4 will always read shortfall on a conducted result and belongs to an external/fuller assessment, not this screen.

D3 — Three items per competence, level-targeted 2 / 3 / 4. Thirty items total, ~10–15 minutes. Difficulty calibration mirrors the Dimension-3 verbs: level-2 items ask the candidate to recognise the basic appropriate action; level-3 items apply a routine, well-defined procedure; level-4 items require independent judgement in a non-routine situation. Count-based mapping (the deployed scorer's contract) is monotone and deterministic.

D4 — Item design discipline. Four options per selected item (the pilot's two-option items were placeholder). One uncontestably-correct key per item — deterministic scoring demands keys that survive challenge. Distractors are behaviourally real (things people actually do), never comic. Prompts ≤ 2 short sentences, plain English, low reading load — the digital screen must not become a covert reading test (the same discipline that put the consent text at Pre-Level-1 readability). Scenarios are Australian VET-learner-natural (course info, learning platform, trainer email, files, phones, public links) and brand-free (no vendor-specific mechanics). No AI-topical items at v1 — DigComp 2.2 carries AI-flagged KSA examples, but AI-surface behaviour churns too fast for a sealed pack; a candidate set for a future version. Item types: selected ×28, task (ordering) ×2; no numeric — the pilot's numeric item was contrived, and forcing a numeric answer onto capability content damages validity (the machinery supports the type; nothing obliges a pack to use it).

D5 — Copy invariants (spec §9.5) applied. No pass/fail or exam language anywhere in prompts or options; the surface is a review, items are questions, the output is a capability level. No item implies a verdict.

D6 — KSA anchors live here, not in the pack JSON. The pack JSON carries exactly the pilot's field set (adding fields to the sealed pack risks the candidate-safe projection contract). Per-item framework anchors and rationale live in this authoring doc, which files in the repo alongside the pack at the publish brief — the mapping provenance is preserved without touching the deployed surface contract.

D7 — Version: digital v2, both environments. Dev holds pilot v1 (pack_235a36e7). Publishing different bytes as "v1" on prod would put two different digests behind one version number across environments — a parity defect. The production pack is v2 everywhere: dev gets v2 alongside the pilot; prod gets v2 as its first pack (a version gap on prod is harmless — UNIQUE (capability_class, version) tolerates gaps, elections seal the pack version at invite, and nothing requires contiguity). The pilot stays dev-only, pilot=1, per commissioning ruling 2.

3. The item set — 30 items, framework-anchored¶

Notation: L2/L3/L4 = the item's level_target. Key = the answer key. Anchor = the DigComp 2.2 Dimension-3 level statement the difficulty mirrors + Dimension-4 KSA example numbers the content operationalises. Correct option is always shown here as (a) for review ease — the publish op randomises option order per item before sealing (see §5 owed items).

1.1 Browsing, searching and filtering data, information and digital content¶

d01 · L2 · selected — "You need to find your training provider's website. What is a reliable way to get there?" (a) Type the provider's name into a search engine and choose its official website · (b) Ask on social media and wait for someone to reply · (c) Type a guess at the address and use whatever page opens · (d) Wait for someone to send you the link. Anchor: L1–2 "find data… through a simple search… find how to access… and navigate"; KSA 6.

d02 · L3 · selected — "You search for a course and the results are mostly about other states. What is the best next step?" (a) Add your state or town to the search words · (b) Open every result one by one until something fits · (c) Use fewer, more general words · (d) Pick the first result anyway. Anchor: L3 "perform well-defined and routine searches"; KSA 7 (improving search results by refining the query).

d03 · L4 · selected — "Your usual search words are not finding what you need for a work task. What is the most effective approach?" (a) Try different words that mean the same thing, and compare what each search returns · (b) Repeat the same search several times · (c) Decide the information does not exist online · (d) Ask one friend and stop looking. Anchor: L4 "organise the searches… non-routine problems / organise personal search strategies"; KSA 10, 11.

1.2 Evaluating data, information and digital content¶

d04 · L2 · selected — "A website says a course guarantees you a job. There is no author, no contact details, and no date. What should you think?" (a) Be cautious — big claims with no source may not be true · (b) It is on the internet, so it has been checked · (c) The professional design means it is trustworthy · (d) A guarantee is a guarantee. Anchor: L1–2 "detect the credibility and reliability of common sources"; KSA 16, 26.

d05 · L3 · selected — "Two websites give different rules about the same first-aid step. What is the best routine check?" (a) Check who published each page and when it was last updated · (b) Trust the longer article · (c) Trust the one with more pictures · (d) Trust whichever came first in the search results. Anchor: L3 "perform the analysis, comparison and evaluation of… well-defined sources"; KSA 18, 23, 26.

d06 · L4 · selected — "A health claim is being shared widely on social media. Before believing it, what is the strongest check?" (a) See whether a trusted organisation (like a government health site) reports the same thing · (b) Count how many likes and shares it has · (c) Read the comments to see what people think · (d) Check whether someone famous shared it. Anchor: L4 "perform the analysis, interpretation and evaluation of data, information and digital content" (non-routine); KSA 16 ("widely reported does not necessarily mean accurate"), 18.

1.3 Managing data, information and digital content¶

d07 · L2 · selected — "You have just finished typing a document you will need again tomorrow. What should you do?" (a) Save it with a name you will recognise · (b) Close it — the computer keeps everything automatically · (c) Take a photo of the screen instead · (d) Leave the computer on with the document open. Anchor: L1–2 "identify how to organise, store and retrieve… in a simple way".

d08 · L3 · selected — "You have many files for your course. What is a routine way to keep them easy to find?" (a) Put them in folders named by subject or week · (b) Keep everything on the desktop · (c) Rename them all 'file1', 'file2', 'file3' · (d) Delete the older ones so the list stays short. Anchor: L3 "select data… in order to organise, store and retrieve them in a routine way / organise them in a routine way in a structured environment".

d09 · L4 · selected — "You need your study files at home, at work, and on your phone. Which choice best fits that need?" (a) Keep them in cloud storage you can sign in to from each device · (b) Save a separate copy on each device and update all three every time · (c) Email the files to yourself after every change · (d) Keep them only on the work computer, where they are safest. Anchor: L4 "organise information, data and content to be easily stored and retrieved" (non-routine); KSA 39 (differentiate storage locations; cloud availability).

2.1 Interacting through digital technologies¶

d10 · L2 · selected — "A form asks for your email address. Which of these is an email address?" (a) jo.smith@example.com · (b) www.josmith.com.au · (c) 0412 345 678 · (d) @josmith. Anchor: L1–2 "select simple digital technologies to interact / identify appropriate simple communication means".

d11 · L3 · selected — "You will miss tomorrow's class, and your trainer needs to know today. Which is the most appropriate choice?" (a) Email or message the trainer directly today · (b) Post about it publicly on your social media · (c) Ask a classmate to mention it if they remember · (d) Explain it at the following week's class. Anchor: L3 "perform well-defined and routine interactions / select well-defined and routine appropriate digital communication means"; KSA 45, 46.

d12 · L4 · selected — "You need a quick answer from your project group. Some members rarely check email, but everyone uses the group chat. The question is not private. Where do you ask?" (a) In the group chat · (b) In separate emails to each member · (c) In a private message to one member only · (d) At the next in-person meeting. Anchor: L4 "select a variety of appropriate digital communication means for a given context" (audience/urgency judgement); KSA 45.

d13 · L2 · selected — "You must send a completed form back to your training provider as a file. Which of these works?" (a) Attach the file to a reply email · (b) Read the answers out over the phone · (c) Post a photo of the form on social media and tag the provider · (d) Describe the form in a text message. Anchor: L1–2 "recognise simple appropriate digital technologies to share… content"; KSA 58.

d14 · L3 · selected — "You used sentences from a website in your assignment. What is the right routine practice?" (a) Say where the words came from — name the source · (b) Change a few words so they count as yours · (c) Nothing — anything on a website is free to copy · (d) Use them only if the site has no copyright notice. Anchor: L3 "illustrate well-defined and routine referencing and attribution practices"; KSA 62, 66.

d15 · L4 · selected — "You are sharing a draft in an online document with two classmates. You do not want anyone else to open it. What do you do?" (a) Share it to their accounts only, rather than turning on 'anyone with the link can open' · (b) Use 'anyone with the link' — nobody else will find it · (c) Let anyone edit, so it is easier for everyone · (d) Post the link in the class-wide forum so your group sees it. Anchor: L4 "manipulate appropriate digital technologies to share… content" (non-routine); KSA 60 (select and restrict with whom content is shared).

2.6 Managing digital identity¶

d16 · L2 · selected — "Before posting a photo publicly, what is a simple way to protect your reputation?" (a) Think about whether you would be comfortable with anyone at all seeing it · (b) Post it now and delete it later if there is a problem · (c) Assume only your friends will ever see it · (d) Post it from a second account instead. Anchor: L1–2 "describe simple ways to protect my reputation online"; KSA 109, 116.

d17 · L3 · selected — "An employer or training provider might look you up online. What is a sensible routine step?" (a) Search your own name to see what appears publicly · (b) Assume nothing about you is online · (c) Ask friends to post more about you · (d) There is nothing anyone can do about it. Anchor: L3 "explain well-defined and routine ways to protect my reputation online / describe well-defined data I routinely produce"; KSA 110 (own-name search, digital footprint).

d18 · L4 · selected — "A website you no longer use still holds your personal details and old posts. What is the most effective action?" (a) Sign in and delete the account, or ask the service to remove your data · (b) Delete the app icon from your phone · (c) Stop thinking about it — old accounts do nothing · (d) Make a new account with a different name. Anchor: L4 "manipulate data I produce through digital tools, environments or services"; KSA 106 (right of erasure), 112 (manage/delete data, deleting unused accounts).

3.1 Developing digital content¶

d19 · L2 · task — "Put these steps in order to write and keep a simple document: 1 — open the writing app · 2 — type your text · 3 — save the file." (a) 1, 2, 3 · (b) 2, 3, 1 · (c) 3, 1, 2 · (d) 2, 1, 3. Anchor: L1–2 "identify ways to create and edit simple content in simple formats"; KSA 118.

d20 · L3 · selected — "Your trainer asks for your work 'as a document file, not a photo'. You have handwritten notes. What do you do?" (a) Type the notes into a document app and save the file · (b) Photograph the notes and rename the image file · (c) Paste the photo into the body of an email · (d) Send a video of yourself reading the notes. Anchor: L3 "indicate ways to create and edit well-defined and routine content in well-defined and routine formats"; KSA 118 (content types and file formats).

d21 · L4 · selected — "You are sending a finished flyer. It must look the same on every device and not be accidentally changed. Which is the best way to send it?" (a) As a PDF copy · (b) As the editable working file · (c) As a link where anyone can edit · (d) As plain text typed into the email body. Anchor: L4 "indicate ways to create and edit content in different formats"; KSA 123 (select the appropriate format for the purpose — editable vs fixed).

4.1 Protecting devices¶

d22 · L2 · selected — "Which is a simple way to help keep your phone or laptop safe to use?" (a) A password or PIN that only you know · (b) Sharing your password with classmates in case you forget it · (c) Turning off software updates so nothing changes · (d) Leaving it unlocked so you can't be locked out. Anchor: L1–2 "identify simple ways to protect my devices… choose simple safety and security measures"; KSA 166, 167.

d23 · L3 · selected — "Your computer says an update with security fixes is available. The good routine practice is to:" (a) Install the update soon · (b) Ignore it — updates slow the computer down · (c) Wait a year to see whether other people like it · (d) Uninstall the program that asked. Anchor: L3 "select well-defined and routine safety and security measures"; KSA 168 (keeping the OS and applications up to date).

d24 · L4 · selected — "You get a text from an unknown number: 'Your parcel is held. Pay $2 here' with a link. What is the safest response?" (a) Do not tap the link; if unsure, check directly with the delivery company yourself · (b) Pay — it is only $2 · (c) Tap the link just to see what it is · (d) Reply asking whether it is real. Anchor: L4 "differentiate risks and threats in digital environments" (non-routine); KSA 170 (scams, identity theft).

4.2 Protecting personal data and privacy¶

d25 · L2 · selected — "A website you have never heard of asks for your bank details 'to verify your identity' before it will show course prices. What should you do?" (a) Do not enter them — seeing prices never needs bank details · (b) Enter them — verification is normal · (c) Enter them, then ask the site to delete them · (d) Enter someone else's details to stay safe yourself. Anchor: L1–2 "select simple ways to protect my personal data and privacy"; KSA 184 (basic security in online payments/details).

d26 · L3 · selected — "An email says your account closes today unless you click a link and re-enter your password. It has spelling mistakes and a strange sender address. The routine safe step is:" (a) Do not click; open the service's real website or app yourself and check there · (b) Click quickly before the account closes · (c) Reply to the email with your password so they can fix it · (d) Click the link but type the password slowly. Anchor: L3 "explain well-defined and routine ways to protect my personal data"; KSA 183 (recognising phishing).

d27 · L4 · selected — "Before you sign up to a new online service with your personal details, what best tells you how your information will be used?" (a) The service's privacy policy or collection notice · (b) The number of downloads it has · (c) Its star rating in the app store · (d) How professional the website looks. Anchor: L4 "discuss ways to protect my personal data… / indicate privacy policy statements of how personal data is used"; KSA 181.

5.1 Solving technical problems¶

d28 · L2 · selected — "A program on your computer is frozen and will not respond. A sensible first step is to:" (a) Close and reopen the program, or restart the computer · (b) Delete the program immediately · (c) Keep clicking the same button many times · (d) Unplug the internet router. Anchor: L1–2 "identify simple technical problems… identify simple solutions"; KSA 217, 224.

d29 · L3 · task — "Your laptop will not open a website. Put these routine checks in a sensible order (least disruptive first): 1 — check the Wi-Fi is connected · 2 — try a different website · 3 — restart the laptop." (a) 1, 2, 3 · (b) 3, 2, 1 · (c) 2, 3, 1 · (d) 3, 1, 2. Anchor: L3 "indicate well-defined and routine technical problems… select well-defined and routine solutions"; KSA 218 (connectivity causes), 224 (step-by-step root-cause). Rubric basis for the key: cheapest/least-disruptive check first, isolate second, restart last — stated in the prompt ("least disruptive first") so the ordering criterion is explicit, not implied.

d30 · L4 · selected — "In an online class, other people cannot hear you. Which is the best way to narrow down the problem?" (a) Check the microphone is not muted in the class app, then test the microphone in your device's settings · (b) Assume the platform is broken and wait · (c) Restart your internet modem · (d) Leave the class and rejoin next week. Anchor: L4 "differentiate technical problems… select solutions" (non-routine); KSA 222 (solving a microphone issue in an online meeting).

4. Scoring — the deterministic level mapping¶

Per competence (all ten identical): count of correct answers → DigComp 2.2 level:

Correct	Level	Meaning (honest floor/ceiling semantics per D2)
0	1	not demonstrated above basic-with-guidance (screen floor)
1	2	foundation, autonomous with guidance available
2	3	intermediate — routine, well-defined problems independently
3	4	intermediate — independent, non-routine (screen ceiling)

level_mapping JSON: every competence key maps {"0":1,"1":2,"2":3,"3":4}. The scorer is the deployed one; nothing changes in code.

5. What the publish op owes (the later brief, Tim's nod required)¶

Tim's item-by-item copy review of raw strings — the G3 copy-checkpoint discipline applies with more force here than it did to consent text: these are the candidate-facing instrument words. Nothing seals until Tim has read every prompt and option as raw strings.
Option-order randomisation before sealing. In this draft the key is always (a) for review ease; the pack must not ship with a positional key pattern. Randomise per item at pack assembly, then the key letters re-map. (The pack is sealed after randomisation — order is then fixed for all candidates under v2, which is what same-published-frame consistency wants.)
~~Cross-check against the repo CSV extraction~~ — DONE in-session 2026-07-10 (§1); the publish brief carries the result, not the task.
Pack assembly + canonical digest via tools/publish-llnd-instrument.mjs (the only writer), pilot=0, version 2, framework_ref 'DigComp 2.2 (JRC128415)' — dev first, gate-walked, then prod on Tim's explicit publish nod.
This authoring doc files in the repo alongside the pack (D6 — the mapping provenance).
Instrument config row (llnd_instrument_config digital v2) sealing framework_ref + item_set_ref + rubric_ref per the deployed convention — confirm the exact ref convention against the CONDUCT-01 artefacts at brief-draft time.
DigComp 3.0 watch item rides unchanged (trigger: DEWR names 3.0 or a 3.0-based ADCF successor → new instrument version; historical screenings stay pinned).

6. Open items for Tim¶

Rule on D1–D7 (or amend — the subset, the level ceiling, item counts, and the v2 numbering are all reversible until publish).
Item-by-item copy review (§5.1). Two items to look at hardest: d29 (the ordering rubric — the prompt now states the criterion, but if "least disruptive first" feels contestable, swap the item for a selected-response troubleshooting item); d10 (near-trivial by design — it is the level-2 floor probe for 2.1; if it reads as condescending, I have a replacement candidate: choosing between email/phone/chat for a simple context).
Whether the LLN-rung brief should draft next session or wait for this pack to publish first.