Test Importer

Drop your authored IELTS Listening/Reading HTML here — we parse TEST_DATA + ANSWERS, keep synced answerData for grading/review, and auto-build an Answer Key view when explanations are already inside the source. Optional Answer HTML is treated as a rich visual layer and checked against the same keys.

document_scanner

NEXUS Listening PDF Transcriber

Convert Gemini/OCR output into parser-ready Listening text with local QA for tables, floor layouts, image alerts, and answer-scope drift.

open_in_new Open Transcriber

1 Source

integration_instructions

Question Frame Prompt (PDF → Parser Text)

Paste into Gemini/Claude with the PDF. Output is strict Listening Plain Text v1 for the Question + transcript frame.

open_in_new Tool

graphic_eq

Timestamp Frame Prompt (Audio → Timestamps + Transcript)

Use in Google AI Studio with audio. Outputs TIMESTAMP_FRAME rows plus word-for-word transcript with `[Qn START - MM:SS]` and `[Qn - MM:SS]` markers.

Audio (mp3)→ AI Studio + this prompt→ Transcript.txt→ feed into HTML / answerHtml builder

school

HTML Builder Prompt — Practice Mode

v5.2 — locked format, anti-drift rules, 4 sequential audio players per-part. Feed Transcribe + Audio Transcript output into an AI along with this prompt → full HTML.

Text + Transcript→ AI + HTML Builder v5.2→ Practice HTML→ drop here

grading

HTML Builder Prompt — Exam Mode

v5.2-EXAM — 1 continuous audio, auto-play, anti-pause, no seek. Use for mock exam bundles. Feed same inputs as Practice but with single audio URL.

Text + Transcript→ AI + HTML Builder v5.2-EXAM→ Exam HTML→ drop here

integration_instructions

Transcribe Prompt (PDF → Text)

Paste into Gemini/Claude along with the PDF. Outputs Passages 1–3 + Questions + Answer Key 1–40, with every accepted variant written explicitly for later alts sync.

PDF / images→ Gemini / Claude + this prompt→ Plain text (3 passages + 40 Q + key)

auto_fix_high

PDF → Reading Parser Frame Prompt

Paste into Gemini/Claude with the PDF when you want the AI to return NEXUS IELTS READING PLAIN TEXT v1 directly. Includes visual/image slots and answer-key QA.

PDF / images→ AI + this prompt→ Reading Plain Text v1→ Plain text tab

terminal

Codex PDF → Reading HTML Prompt

For Codex: attach/source the PDF, learn the locked Practice Mode prompt, then produce one importer-ready IELTS Reading HTML with TEST_DATA + ANSWERS.c/l/v/p + explicit alts.

PDF / source→ Codex + this prompt→ Importer-ready HTML→ drop here

school

HTML Builder Prompt — Practice Mode

v5.1 LOCKED FORMAT — 3-passage split layout, five-phase accent highlights, importer-ready one-file output. Feed transcribed text to an AI + this prompt → full HTML with TEST_DATA + ANSWERS.c/l/v/p + explicit alts.

Transcribed text→ AI + HTML Builder v5.1→ Synced Practice HTML→ drop here

rule_settings

NEXUS THPT Plain Text Parser Standard

For THPT / Grade 10, paste the AI's 5-block plain text in the Plain text tab. The parser reads dynamic IMAGE ALERT lines, creates labelled image upload slots by Q/range, preserves explanations, and saves clean testData v2. Copy Gemini and Copy Codex now include the TO G10 quality checklist for type-specific explanations, IPA/stress checks, dictionary no-reveal rules, answer variants, section points, and rewrite scope.

open_in_new Open standard

Copy Gemini/Codex prompt→ AI returns 5-block plain text→ Paste Plain text→ Upload IMAGE ALERT slots→ Save testData v2→ Practice / Make Lecture

code_blocks

Test HTML

Your teacher-authored file — drop or click. Must contain TEST_DATA + ANSWERS.c/l/v/p; accepted variants belong in alts.

menu_book

Answer Key HTML (optional)

Optional rich explanation file — drop or click. Its keys are synced against testData; mismatches are flagged before saving.

…or paste HTML source directly

Parses locally — no API call.

description

Question Frame Prompt (PDF → Parser Text)

Same prompt as the main frame. Paste into Gemini/Claude with PDF to get strict parser-ready question layout + answer key.

open_in_new Tool

graphic_eq

Audio Transcript Prompt (timestamps + transcript)

Google AI Studio + audio → transcript with dual Q markers. Paste the marker rows into the timestamp frame for precise review replay and highlights.

data_object

JSON Builder Prompt (Text → JSON)

Paste this + the transcribed text into any AI. Output = pure JSON matching the Test Importer schema. Paste back into the textarea below.

Transcribed text→ Any AI + JSON Prompt→ JSON→ paste below

auto_fix_high

Reading PDF/Text → Parser Frame

Phase 2.1 workspace: copy a PDF-to-text prompt, or paste old/OCR text and normalize it into NEXUS IELTS READING PLAIN TEXT v1. This prepares the source for the real Reading parser in Phase 2.2.

open_in_new Open spec

PDF/images→ PDF to text→ Normalize→ Reading Plain Text v1

edit_note

AI Text → Writing Lecture + Test

Use clear labels such as TITLE:, LECTURE:, and TEST:. If there is no LECTURE: block, NEXUS treats it as Writing Test only and keeps the test dimmed until START.

record_voice_over

AI Text → Speaking Lecture + Practice

Copy the v6.2 prompt for full lessons, or use the AI Studio audio prompt when you already have examiner audio and need timestamps for audio-only Part 1/3 exam delivery.

open_in_new Audio spec open_in_new Open spec

graphic_eq

AI Text → IELTS Listening Test

Paste the question/transcript frame, then optionally paste standalone timestamp and explanation frames. NEXUS merges them by Q number so timing fixes do not disturb the test layout.

open_in_new Open spec

Questions + transcript→ Timestamp frame→ Explanation frame→ parse + preview

looks_one Task 1 AI source

chart / table / map / process

add_photo_alternate

Drop Task 1 image here

AI-downloaded images are auto-detected, resized, compressed, and converted before save.

visibility_lock Chart Transcriber source teacher/GPT only

NEXUS merges this into Task 1 during Parse, stores it as hidden marker context, and keeps it out of the student test flow.

looks_two Task 2 AI source

essay prompt / model / feedback

Audio parserSplit input into Part 1 audio timestamp text, Part 2 cue card text, and Part 3 audio timestamp text.

graphic_eq IELTS Listening

IELTS Listening parser workspace

Paste the required question frame first, then add timestamp and explanation frames only when they are ready.

1 Question 2 Timestamp 3 Explanation

format_list_bulleted Question + transcript frame

required

Copy prompt for this frame

Use with the IELTS Listening PDF/images. The AI must return one strict parser-ready source for this main box.

open_in_new Open

Paste output here: === NEXUS IELTS LISTENING PLAIN TEXT v1 === META / AUDIO / VISUAL_ALERTS / PART 1-4 / ANSWER_DATA / TRANSCRIPT / QA_FLAGS === END NEXUS IELTS LISTENING PLAIN TEXT v1 ===

Question transcript sourcemain frame

This frame owns the test layout: META/AUDIO/PART blocks, questions, options, answer key, visual alerts, and optional transcript.

timer Timestamp frame

optional override

Copy prompt for this frame

Use in Google AI Studio with audio + answer key. Output gives replay start and answer timestamp per Q.

open_in_new Open

Paste output here: TIMESTAMP_FRAME Q1 | start=02:55 | answer=02:58 | evidence=... END_TIMESTAMP_FRAME Full timestamped transcript is also accepted here.

Replay timestamp sourceaudio frame

START becomes replay start; answer timestamp becomes the Q badge/highlight point. This frame wins over inferred transcript timing.

psychology_alt Detailed explanation frame

optional review layer

Copy prompt for this frame

Use after the question frame and timestamp frame are ready. Output only rich review explanations by Q number.

open_in_new Open

Paste output here: EXPLANATION_FRAME Q1 Answer: ... Keyword: ... Evidence: ... Explanation: ... Paraphrase: ... END_EXPLANATION_FRAME

Review explanation sourceanswerData layer

Explanation rows merge into answerData for submitted review. Answer conflicts are surfaced by QA after parsing.

menu_book IELTS Reading

Reading transcriber and normalizer

Paste raw PDF/OCR text or the output from the PDF-to-text prompt. NEXUS will normalize passages, question blocks, image alerts, and answer key into the canonical parser frame.

description Raw Reading source

paste here

Copy prompt for PDF to text

Use this when you want a clean human transcription first, then paste it below for normalization.

open_in_new Open

Raw PDF/OCR/text sourceinput

rule_settings Normalized parser frame

phase 2.1

Direct PDF to parser frame

Use this prompt when the AI should return NEXUS IELTS READING PLAIN TEXT v1 directly.

open_in_new Open

0 passages

0 answers

0 visuals

Normalize a Reading source to see blockers, warnings, and parser-readiness checks.

Reading Plain Text v1output

draft

NEXUS General Practice — Plain Text Builder

Two flavours: Gemini (full inline prompt — paste into a Gem or chat) or GPT (lean < 8K system instruction that fetches the full spec via web browsing). Both run a Pre-flight Review, wait for your confirmation, then emit the 5-block plain text. Paste the AI's output below and hit Parse.

open_in_new Gemini (full) smart_toy GPT (lean)

Raw exam content→ Gemini / GPT + system instruction→ 5-block plain text→ paste below

rule_settings

THPT Plain Text — Gemini Prompt

Copy the direct Gemini prompt for parser-ready 5-block THPT text. PDF không có đáp án vẫn được xử lý theo flow transcribe + tự giải + giải thích chi tiết bằng tiếng Việt. Pronunciation requires UnderlinedPart plus option-level markers like play[s] / work[ed].

open_in_new Open prompt

PDF / image / OCR→ Gemini prompt→ 5-block plain text→ paste below

Available for IELTS Reading, IELTS Listening, IELTS Writing, IELTS Speaking, and General Practice.