Resolve pending commit links of the MIL-009 plan

MIL-009 plan and documents: executable with PyInstaller
Milestone MIL-009, story US-001.07, ADR-0004, and the amended BC-001, DCD-001 and PP-001. Review records RC-042 to RC-047, all Go (autonomous mode, not independent). Task: MIL-009#1
2026-10-15 00:17:31 +08:00 · 2026-10-15 00:17:27 +08:00 · 2026-10-14 17:58:52 +02:00 · 2026-10-14 23:58:46 +08:00 · 2026-10-14 23:58:31 +08:00 · 2026-10-14 17:45:16 +02:00
52 changed files with 2055 additions and 122 deletions
@@ -15,7 +15,7 @@ A flash card program for learning the most frequent Malay words with their Danis

 ## 🧭 Overview

-The program shows a Malay word on a card. After three seconds the card turns over and shows the Danish translation. You press the green tick when you know the word and the red cross when you do not. A known word is removed from the words to learn, and the words still to learn are saved, so the next start continues where you stopped.
+The program first asks for a level: the first 100 words, the first 1000 words or all words of the word list, which is ordered from the most to the least frequent word. Then it shows a Malay word on a card. After three seconds the card turns over and shows the Danish translation. You press the green tick when you know the word and the red cross when you do not. A known word is removed from the words to learn, and the words still to learn are saved, so the next start continues where you stopped.

 | Part | What it is for |
 | --- | --- |
@@ -30,9 +30,9 @@ The core program uses only the Python standard library. Spoken Malay words are a

 ### Screenshots

-| Front: the Malay word | Back: the Danish translation after 3 seconds |
-| --- | --- |
-| ![The front of a card][screenshot-front] | ![The back of a card][screenshot-back] |
+| Choose your level | Front: the Malay word | Back: the Danish translation after 3 seconds |
+| --- | --- | --- |
+| ![The level screen][screenshot-level] | ![The front of a card][screenshot-front] | ![The back of a card][screenshot-back] |

 ### Where is each assignment step?

@@ -42,6 +42,7 @@ The core program uses only the Python standard library. Spoken Malay words are a
 | 2. New flash cards | `CsvVocabularyRepository` (the standard library `csv` instead of `pandas`), `Deck.draw`, `StudySession.next_card`, `FlashCardView.next_card` |
 | 3. Flip the cards | `FlashCardView.flip_card` and `FlashCardView.flip_timer` (`after`, `after_cancel`), `FlashCardViewModel.flip_card` |
 | 4. Save your progress | `FlashCardView.is_known`, `StudySession.mark_known`, `CsvProgressRepository` |
+| Extra: level choice | `LevelView`, `LevelViewModel`, `Level`, `Deck.within`, `StudySession.next_card`, `open_study_session` |
 | Extra: spoken word | `GttsSpeech`, `NullSpeech`, `build_speech` in `presentation/app.py` |

 The wiki ([English][wiki-en], [Dansk][wiki-da], [Bahasa Malaysia][wiki-ms]) explains the same in three languages.
@@ -92,14 +93,15 @@ python -m pip install -e ".[dev]"
 python -m flash_card
 ```

-The program is run from the activated `.venv`. The window opens with a card. Press the green tick for a word you know and the red cross for a word you do not. Options:
+The program is run from the activated `.venv`. The window opens with the level screen: choose 100 words, 1000 words or All words. Then a card appears. Press the green tick for a word you know and the red cross for a word you do not. Options:

 | Option | Meaning |
 | --- | --- |
 | `--data-dir <folder>` | Folder with `ms-da.csv` and `words_to_learn.csv` |
+| `--level {100,1000,all}` | Start at that level and skip the level screen |
 | `--no-speech` | Do not speak the Malay word |

-To start again, delete `data/words_to_learn.csv`.
+Known words stay known at every level: the level only limits which of the words you do not know yet come up. To start again, delete `data/words_to_learn.csv`.

 ## 🧪 Tests

@@ -149,6 +151,7 @@ The documents a newcomer needs, in reading order.
 ---

 [license]: ./LICENSE
+[screenshot-level]: ./docs/images/screenshot-level.png
 [screenshot-front]: ./docs/images/screenshot-front.png
 [screenshot-back]: ./docs/images/screenshot-back.png
 [wiki-en]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/wiki/Home
@@ -0,0 +1,55 @@
+# ADR-0004: Windows executable built with PyInstaller
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | ADR-0004 |
+| CrossReference | [BC-001], [US-001], [DCD-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+Allowed Status values: `Proposed`, `Accepted`, `Rejected`, `Deprecated`, `Superseded by ADR-NNNN`.
+
+---
+
+## Context
+
+The Product Owner (S01) wants an executable file, so that the program can be started without a Python installation (US-001.07). The program has images, a word list, a progress file that must survive restarts, and an optional speech extra. Options:
+
+1. A one-directory build (a folder with the executable and its libraries): starts faster and the data can sit beside the files, but it is a folder of hundreds of files to hand over.
+2. A one-file build (a single `flash-card.exe` that unpacks itself to a temporary folder): one file to hand over, slower start; resources are in a temporary folder, so progress needs a place beside the executable.
+3. Another packager (Nuitka, cx_Freeze) or an installer: more tooling and a larger change for the same result.
+
+## Decision
+
+Build a one-file, windowed (no console) Windows executable with PyInstaller by `tools/build_executable.py`. The images and the word list are inside the executable. The progress file and an optional word list are in a `data/` folder beside the executable (found with `sys.executable` when `sys.frozen` is set), and `--data-dir` still overrides it. PyInstaller is only in the optional extra `build`, so `dependencies` stays empty (ADR-0001). The executable is a build output and is not committed.
+
+## Consequences
+
+**Positive:**
+
+- One file to hand over, and no Python, `.venv` or pip needed to run it.
+- The source run is unchanged; only the path rule differs when frozen, and it is tested without PyInstaller.
+- The word list stays editable beside the executable.
+
+**Negative:**
+
+- PyInstaller cannot cross-compile: the executable must be built on Windows (and on Linux or macOS for those systems).
+- A one-file executable unpacks to a temporary folder at each start and is larger with the speech packages; unsigned executables can trigger Windows SmartScreen or antivirus warnings.
+- The Windows Store Python can cause Tcl/Tk path problems with PyInstaller; a build with it must be checked by a real start.
+
+## Affected Artifacts
+
+- [BC-001] — scope: the executable is in scope; installer, signing and release upload are out
+- [US-001] — US-001.07 Start without Python
+- [DCD-001] — packaging support and the choice of the word list file in the composition root
+
+---
+
+[BC-001]: ../business-case.md
+[US-001]: ../user-stories.md
+[DCD-001]: ../dcd.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -14,7 +14,7 @@ document of a type. `Primary File` may contain a glob (e.g.
 | BC | Business Case | docs/business-case.md | 002 |
 | SA | Stakeholder Analysis | docs/stakeholder-analysis.md | 002 |
 | PP | Project Plan | docs/project-plan.md | 002 |
-| MIL | Milestone / Gateway | docs/milestones/*.md | 009 |
+| MIL | Milestone / Gateway | docs/milestones/*.md | 010 |
 | UCD | Use Case Diagram | docs/use-case-diagram.md | 002 |
 | US | User Story | docs/user-stories.md | 002 |
 | UC | Use Case | docs/uc-*/uc.md | 002 |
@@ -24,8 +24,8 @@ document of a type. `Primary File` may contain a glob (e.g.
 | SSD | System Sequence Diagram | docs/uc-*/ssd.md | 002 |
 | DCD | Design Class Diagram | docs/dcd.md | 002 |
 | DICT | Domain Dictionary | docs/dictionary.md | 002 |
-| ADR | Architecture Decision Record | docs/adr/adr-NNNN-*.md | 0004 |
-| RC | SQA Review Record | docs/sqa/reviews/rc-*.md | 041 |
+| ADR | Architecture Decision Record | docs/adr/adr-NNNN-*.md | 0005 |
+| RC | SQA Review Record | docs/sqa/reviews/rc-*.md | 048 |
 | TM | Traceability Matrix | docs/sqa/traceability-matrix.md | 002 |

 ## Languages
@@ -12,7 +12,8 @@
 | Date | Status | Author | Reviewer | Change | Commit |
 | --- | --- | --- | --- | --- | --- |
 | 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [c3fbcbc] |
-| 2026-10-14 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | In Scope: choice of level (new request of S01) | [0b983b7] |
+| 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | In Scope: choice of level (new request of S01) | [0b983b7] |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Scope: executable built with PyInstaller (new request of S01) | [768ca34] |

 ---

@@ -54,6 +55,7 @@ A small, well-structured reference solution gives the Product Owner a language l
 - Saving the words still to learn to `data/words_to_learn.csv` and using that file on the next start.
 - Optional spoken Malay word (text to speech) as an extra that can be installed.
 - A choice of level (the first 100 words, the first 1000 words or all words of the frequency ordered word list) before learning starts.
+- A single Windows executable file, built with PyInstaller by a script of this repository, that runs without an installed Python.
 - `pyproject.toml`, `.gitignore`, `Doxyfile`, virtual environment instructions, `pytest` tests and `constants.py`.
 - README from the framework template, screenshots, and a wiki in English, Danish and Bahasa Malaysia.
 - Repository description and topics on the git host.
@@ -63,7 +65,7 @@ A small, well-structured reference solution gives the Product Owner a language l

 - Other languages: the user interface labels are Malay and Danish only; another word list works by changing the data file but is not delivered.
 - Spaced repetition scheduling, statistics, accounts, cloud synchronisation and mobile apps.
- Packaging for distribution on PyPI or as an installer.
+- Packaging for distribution on PyPI, an installer, a signed executable and an upload of the executable to a release (the executable built with PyInstaller is in scope; the file itself is not committed).
 - Redistribution of the course's own image files; the card and button images are generated for this project.
 - A translation quality review of the word list.

@@ -150,3 +152,4 @@ Proceed — the scope is small, the risks are mitigated, and the result serves t
 [SA-001]: ./stakeholder-analysis.md
 [c3fbcbc]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/c3fbcbcfaf799e0ae34aed7e14d3bc8a62f03db3
 [0b983b7]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/0b983b7fdadf6958717c1d0227d0f2ebe037f7b1
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -4,19 +4,20 @@
 | Key | Value |
 | --- | --- |
 | ID | DCD-001 |
-| CrossReference | [DM-001], [SD-001], [DICT-001], [ADR-0002] |
+| CrossReference | [DM-001], [SD-001], [DICT-001], [ADR-0002], [ADR-0004] |

 ## Version History
 | Date | Status | Author | Reviewer | Change | Commit |
 | --- | --- | --- | --- | --- | --- |
 | 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [c3fbcbc] |
-| 2026-10-14 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added Level, Deck.within, the level argument and the level screen | [0b983b7] |
+| 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added Level, Deck.within, the level argument and the level screen | [0b983b7] |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Scope note: runtime path rules and the executable build are packaging support, not design classes | [768ca34] |

 ---

 ## Purpose and Scope

-The design classes of the program in `src/flash_card/`, in four layers (domain, application, infrastructure, presentation) with the presentation pattern MVVM ([ADR-0002]). The classes refine the concepts of [DM-001]; names are the IT terms of [DICT-001]. Methods trace to the operation contracts of [OC-001] and the sequence diagrams of [SD-001] (the SSD messages are in [SSD-001]).
+The design classes of the program in `src/flash_card/`, in four layers (domain, application, infrastructure, presentation) with the presentation pattern MVVM ([ADR-0002]). The classes refine the concepts of [DM-001]; names are the IT terms of [DICT-001]. Methods trace to the operation contracts of [OC-001] and the sequence diagrams of [SD-001] (the SSD messages are in [SSD-001]). The module `flash_card/runtime.py` (where the resources and the data folder are when the program is frozen by PyInstaller) and the scripts in `tools/` are packaging support ([ADR-0004]): they have no system operation and are not design classes; `app.main` is still the only place that names the adapters, and it chooses the word list file.

 ## Diagram

@@ -208,7 +209,7 @@ Main ..> FlashCardView : composes
 | Ports and adapters | `SpeechPort`, `GttsSpeech`, `NullSpeech` | Speech is an optional extra; the core runs without it |
 | Null Object | `NullSpeech` | No `if speech:` checks in the use case |
 | MVVM | `FlashCardView`, `FlashCardViewModel`, `LevelView`, `LevelViewModel` | Behaviour is tested without Tkinter ([ADR-0002]) |
-| Composition root | `app.main` | The one place that names the adapters and shows the level screen |
+| Composition root | `app.main` | The one place that names the adapters, shows the level screen and chooses the word list file (beside the executable, else inside it) |
 | Value object | `Level`, `Card`, `LanguagePair` | Immutable data with its own rule (`Level.first`) |

 ## Dependency Check
@@ -223,5 +224,7 @@ Imports point inward: `domain` imports nothing from the other packages; `applica
 [OC-001]: ./uc-001/oc.md
 [SD-001]: ./uc-001/sd.md
 [ADR-0002]: ./adr/adr-0002-clean-architecture-and-mvvm.md
+[ADR-0004]: ./adr/adr-0004-pyinstaller-executable.md
 [c3fbcbc]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/c3fbcbcfaf799e0ae34aed7e14d3bc8a62f03db3
 [0b983b7]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/0b983b7fdadf6958717c1d0227d0f2ebe037f7b1
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,83 @@
+# MIL-009: Executable with PyInstaller
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | MIL-009 |
+| CrossReference | [BC-001], [US-001], [UC-001], [DCD-001] |
+| Language | en |
+| Domain | it |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Purpose
+
+Decide whether the Learner can start the program from one executable file without a Python installation: the program is packaged with PyInstaller by a script of this repository, and behaves in the executable as it does from the source.
+
+## Deliverable
+
+`python tools/build_executable.py` builds `dist/flash-card.exe` (one file, no console window, an icon, the images and the word list inside, the speech extra included). The executable shows the level screen, keeps the progress in `data/words_to_learn.csv` next to it and prefers a `data/ms-da.csv` next to it over the word list inside. The executable itself is not committed to git; README and wiki (English, Danish, Bahasa Malaysia) explain how to build and use it.
+
+## Go / No-Go Criteria
+
+| # | Criterion (objectively checkable) | Go | No-Go |
+| --- | --- | --- | --- |
+| 1 | `python tools/build_executable.py` ends with exit code 0 and produces `dist/flash-card.exe` | File exists and is larger than 1 MB | No file or a build error |
+| 2 | The executable starts on this machine with the virtual environment not activated and shows the level screen without a console window | A window titled "Flashy" exists after 5 seconds and the process is a Windows GUI program | The process ends or shows a console |
+| 3 | With a `data/ms-da.csv` copied next to the executable it uses that file; without it the word list inside the executable is used | Smoke test with a marked list of two words, then without | The wrong list is used |
+| 4 | Marking a word known in the executable writes `data/words_to_learn.csv` next to it and the next start continues from it | Smoke test reads the file after the run | No file or progress lost |
+| 5 | The four images and the icon are inside the executable and are found when frozen | Card and button images are visible in a captured screenshot of the executable | A missing image or an exception |
+| 6 | Path resolution when frozen (`sys.frozen`, `sys._MEIPASS`) is tested without PyInstaller | Unit tests pass | Not covered |
+| 7 | `dependencies` in `pyproject.toml` is still empty; PyInstaller is only in the extra `build` | File check | A runtime dependency was added |
+| 8 | BC-001, DCD-001 and ADR-0004 are updated and reviewed `Go`; README and wiki in three languages describe the build | Review records `Go`; `check-readme.sh` exit 0 | A document is missing |
+| 9 | Code review against QC-PY-001 has no failed Mandatory criterion; `pytest`, `ruff`, `mypy --strict`, `lint-imports` and `doxygen` are clean | `RC-*` verdict Go and exit codes 0 | Otherwise |
+
+## Dependencies
+
+| Depends on | Reason |
+| --- | --- |
+| MIL-008 | The finished program, with the level screen, that is packaged |
+
+## Traceability
+
+| Business Case objective / KPI / user story | Reference |
+| --- | --- |
+| Objective 4 (a repository others can use), extended by the new request of S01; success criterion 4 (time to run) | [BC-001] |
+| US-001.07 Start without Python | [US-001] |
+
+## Ownership
+
+| Role | Stakeholder ID (SA) |
+| --- | --- |
+| Owner | S01 |
+| Approving reviewer | S01 (AI as stand-in in autonomous mode) |
+
+## Target Date
+
+2026-10-16 — within the delivery window that ends 2026-10-21 ([BC-001] constraint).
+
+## Tasks
+
+| # | Task | Summary | Needs its own Use Case/User Story? | Reference |
+| --- | --- | --- | --- | --- |
+| 1 | Update the documents for the executable | Amend BC-001 (In Scope: an executable built with PyInstaller; Out of Scope: an installer and a signed release stay out), add ADR-0004 (PyInstaller one-file windowed build, word list inside and a `data` folder beside the executable for progress, Windows only), and add the runtime path resolution to DCD-001. Every changed document gets a Version History row and a review. | No | BC-001, DCD-001 |
+| 2 | Resolve resource and data paths when frozen | The images are found under `sys._MEIPASS` and the default data folder is `data/` next to the executable when `sys.frozen` is set; the source run keeps its paths. The rule is one function used by `constants.py`, tested by simulating the frozen state. | Yes | US-001.07 |
+| 3 | Prefer the word list next to the executable | When the data folder has no `ms-da.csv` the program uses the copy inside the executable; the progress file is always written to the data folder; `--data-dir` still overrides both. | Yes | US-001.07 |
+| 4 | Add the build script and the build extra | `tools/build_executable.py` calls PyInstaller (one file, windowed, name `flash-card`, the assets and the word list as data, the icon, the speech packages as hidden imports); `pyproject.toml` gets the extra `build`; `build/`, `dist/` and `*.spec` are ignored. | No | ADR-0004 |
+| 5 | Generate the application icon | `tools/generate_images.py` also writes `flash_card.ico` (the card with the tick) from the same drawing code, so no external asset is used. | No | |
+| 6 | Build and smoke-test the executable | Build it, start it with the virtual environment not active, check the window, the images, the level choice, the progress file next to it, the word list override and the speech import; record the result in the review record. | Yes | US-001.07 |
+| 7 | Test the path rules and the build arguments | Unit tests for the frozen and source path rules and for the argument list the build script gives PyInstaller (without running PyInstaller). | No | |
+| 8 | Update README and wiki | README and the Install page of the wiki in English, Danish and Bahasa Malaysia: how to build, where the executable and the data folder are, that Python is not needed to run it. The wiki is republished after the merge. | No | |
+
+---
+
+[BC-001]: ../business-case.md
+[US-001]: ../user-stories.md
+[UC-001]: ../uc-001/uc.md
+[DCD-001]: ../dcd.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -4,7 +4,7 @@
 | Key | Value |
 | --- | --- |
 | ID | PP-001 |
-| CrossReference | [BC-001], [SA-001], [US-001], [MIL-001], [MIL-002], [MIL-003], [MIL-004], [MIL-005], [MIL-006], [MIL-007], [MIL-008] |
+| CrossReference | [BC-001], [SA-001], [US-001], [MIL-001], [MIL-002], [MIL-003], [MIL-004], [MIL-005], [MIL-006], [MIL-007], [MIL-008], [MIL-009] |
 | Language | en |
 | Domain | it |

@@ -12,7 +12,8 @@
 | Date | Status | Author | Reviewer | Change | Commit |
 | --- | --- | --- | --- | --- | --- |
 | 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [c3fbcbc] |
-| 2026-10-14 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added gateway MIL-008 Choose level (new request of S01) | [0b983b7] |
+| 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added gateway MIL-008 Choose level (new request of S01) | [0b983b7] |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added gateway MIL-009 Executable with PyInstaller (new request of S01) | [768ca34] |

 ---

@@ -40,6 +41,7 @@ This plan schedules the seven phases (gateways) that deliver the flash card prog
 | Spoken Malay word | [MIL-006] | 2026-10-19 | 2026-10-19 | S01 | US-001.05 | Text to speech as optional extra | [Milestone MIL-006][MS-006] |
 | Documentation and publication | [MIL-007] | 2026-10-19 to 2026-10-21 | 2026-10-21 | S01 | none | README, screenshots, wiki in 3 languages, description and topics | [Milestone MIL-007][MS-007] |
 | Choose level | [MIL-008] | 2026-10-20 | 2026-10-20 | S01 | US-001.06 | A level choice (100, 1000, all words) before the cards; documents, wiki and screenshots updated | |
+| Executable with PyInstaller | [MIL-009] | 2026-10-16 | 2026-10-16 | S01 | US-001.07 | One Windows executable that runs without Python, built by a script; documents, wiki and README updated | |

 ```plantuml
@startgantt
@@ -60,6 +62,8 @@ Project starts 2026-10-14
 [Documentation Go/No-Go] happens 2026-10-21
 [Choose level] starts 2026-10-20 and ends 2026-10-20
 [Choose level Go/No-Go] happens 2026-10-20
+[Executable] starts 2026-10-16 and ends 2026-10-16
+[Executable Go/No-Go] happens 2026-10-16
@endgantt
 ```

@@ -75,12 +79,13 @@ Project starts 2026-10-14
 | README, screenshots, wiki in three languages | MIL-007 |
 | Repository description and topics | MIL-007 |
 | Choice of level (new request after MIL-007) | MIL-008 |
+| Executable built with PyInstaller (new request after MIL-008) | MIL-009 |
 | Framework documents of the baseline | Baseline pull request before MIL-001 |

 ## Dependencies

 ```
-Baseline → MIL-001 → MIL-002 → MIL-003 → MIL-004 → MIL-005 → MIL-006 → MIL-007 → MIL-008
+Baseline → MIL-001 → MIL-002 → MIL-003 → MIL-004 → MIL-005 → MIL-006 → MIL-007 → MIL-008 → MIL-009
 ```

 A No-Go on a phase moves every later date by the days the rework takes; MIL-008 was added after MIL-007 and takes the unplanned day 2026-10-20, so no buffer is left inside the window; a No-Go moves the end date past 2026-10-21 and S01 decides whether to accept that.
@@ -111,6 +116,7 @@ A No-Go on a phase moves every later date by the days the rework takes; MIL-008
 [MIL-006]: ./milestones/mil-006-malay-speech.md
 [MIL-007]: ./milestones/mil-007-documentation-and-publication.md
 [MIL-008]: ./milestones/mil-008-choose-level.md
+[MIL-009]: ./milestones/mil-009-executable.md
 [c3fbcbc]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/c3fbcbcfaf799e0ae34aed7e14d3bc8a62f03db3
 [MS-001]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/milestone/165
 [MS-002]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/milestone/166
@@ -120,3 +126,4 @@ A No-Go on a phase moves every later date by the days the rework takes; MIL-008
 [MS-006]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/milestone/170
 [MS-007]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/milestone/171
 [0b983b7]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/0b983b7fdadf6958717c1d0227d0f2ebe037f7b1
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,64 @@
+# RC-041: Review of PY-MIL-008
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-041 |
+| CrossReference | [PY-MIL-008], [QC-PY-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-14 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [f45d603] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [PY-MIL-008]
+- Checklist used: [QC-PY-001]
+- Scope: full review
+- Language and domain: n/a (technical type)
+- Language reviewer: none
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Checklist Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | Packages, modules, functions, variables, classes and constants follow PEP 8 casing (`snake_case`, `PascalCase`, `UPPER_SNAKE`) | Pass | PEP 8 casing holds in all new modules (`level.py`, `level_view.py`, `window.py`); ruff rule set N passes. |
+| 2 | Names state purpose in the domain's language; no unexplained abbreviations, no single-letter names outside tiny scopes | Pass | Names state purpose and are the dictionary's IT terms (`Level`, `Deck.within`, `LevelViewModel`, `LevelChoice`, `Application.start`). |
+| 3 | Code is produced by the project's formatter and passes its linter with no unexplained suppressions | Pass | `ruff check` and `ruff format --check` pass on 67 files with no suppressions. |
+| 4 | Every function and method signature is type-annotated, including `-> None` | Pass | Every signature is annotated; `mypy --strict` passes on 44 source files and on `tools/capture_screenshots.py`. |
+| 5 | No bare `except:`, no swallowed exceptions; specific exceptions are raised and the cause is kept (`raise ... from`) | Pass | `Level` raises `ValueError` for a limit below 1; a failed session start after a click is logged, closes the window with exit code 1 and never swallows silently; no bare `except`. |
+| 6 | No mutable default arguments and no shadowed builtins | Pass | No mutable defaults, no builtin shadowed. |
+| 7 | Files, locks and connections are managed with context managers | Pass | No new file handling beyond the existing context managers. |
+| 8 | Public modules, classes and functions have docstrings that say what, not how | Pass | New and changed members have Doxygen docstrings; `doxygen Doxyfile` ends with 0 warnings. |
+| 9 | Logging uses `logging`, not `print`; no secrets or personal data in log output | Pass | No `print`; failures are logged with `logging`; no personal data is logged. |
+| 10 | Classes and operations trace to the Design Class Diagram they implement; deviations are recorded | Pass | `Level`, `Deck.within`, `Deck.draw(among)`, `StudySession(level_cards)`, `open_study_session(..., level)`, `LevelChoice`, `LevelViewModel`, `LevelView` and `Application` match the revised DCD-001, OC-001 `chooseLevel` and SD-001. Deviations: `Application` (root, repositories, speech) with `start`, `show_level_screen` and `exit_code` is the composition root detail that DCD-001 shows as `app.main`; `StudySession.remaining_in_level` and `LevelView.frame` are additions. |
+| 11 | Tests exist for new behaviour, are named for the behaviour, and do not depend on order or the network | Pass | 191 tests (3 consecutive full runs green) cover `Level`, `Deck.within`, the level filter, the cross-level progress rules on the real word list, the view model, `--level` and the real level window flow; none uses the network. |
+| 12 | Type checker runs in strict mode without errors; `Any` is justified in a comment | Pass | `mypy --strict` reports no issue; no `Any`. |
+| 13 | Dependencies are declared and pinned in the project's dependency file, none unused | Pass | `dependencies` stays empty; no new package. |
+| 14 | Each piece of knowledge (a business rule, constant, format, validation or query) is defined in one place; code is merged only where it expresses the same knowledge, not where it merely looks alike | Pass | The level rule exists only in `Level.first` and `Deck.within`; the choices, labels and sizes exist once in `constants.py`; both views use `configure_window`. |
+| 15 | Imports point inward: business-rule modules import no framework, database, UI or delivery-mechanism code, and there are no import cycles | Pass | `lint-imports` reports `Layers point inward` kept; a grep finds no `tkinter`, `csv` or `random` import in `domain` or `application`; the level view model imports no Tkinter. |
+| 16 | SOLID holds: a class or module has one reason to change; new behaviour is added by extension, not by editing a type test in several places; subclasses honour the contract of their base; protocols are narrow; high-level code depends on protocols or abstract types and receives its implementations | Pass | The view model owns the choices and the view draws them; the composition root only wires; the session depends on protocols and receives its collaborators. |
+| 17 | Source is laid out in the layers and folders of the clean architecture folder structure (`domain`, `application`, `infrastructure`, `presentation` packages under `src/`, and `unit`, `integration` and `e2e` under `tests/`), in `snake_case`, and imports between the layers point inward | Pass | `presentation/features/level/` and `presentation/shared/` follow the clean-architecture folder structure; tests stay in `unit`, `integration`, `e2e` (with a session-scoped `tk_root` in `tests/e2e/conftest.py`). |
+| 18 | File and folder names are English, contain no space and no character outside letters, digits, `-`, `_` and `.`, and follow the casing of the language for files | Pass | File and folder names follow the rule. |
+
+## Overall Verdict
+
+Go — every Mandatory criterion passes. Reviewed by AI as stand-in for S01; not independent of the author.
+
+Milestone MIL-008 Go criteria: 1 level 100 draws exactly the first 100 rows over a full run of the real list (integration test); 2 levels 1000 and All use all 1000 words, and a shorter list is used whole (unit and integration tests); 3 words known at level 100 stay known at All and the progress file keeps every unknown word, also outside the level (integration test); 4 a level with nothing left gives the finished message and no failure (unit and e2e tests); 5 the level view model has no Tkinter import and its tests pass without a display; 6 `--level 100|1000|all` skips the screen and an unknown value exits with 2 (unit tests); 7 UC-001, DM-001, DICT-001, SSD-001, OC-001, SD-001, DCD-001, BC-001, US-001, PP-001 and MIL-008 were re-reviewed `Go` (RC-030 to RC-040), the wireframe has Screen 1, and the README and the wiki in en, da and ms describe the level; 8 this record, `pytest` 191 passed, ruff, mypy, lint-imports and doxygen clean. A real screenshot of the level screen was captured and inspected after fixing a too-small window. The Danish and Malay wiki additions were written by the AI and are not read by a native speaker. All Go.
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| none | - | - |
+
+---
+
+[PY-MIL-008]: ../../../src/flash_card
+[QC-PY-001]: ../../../framework/qc/code/qc-programming-python.md
+[f45d603]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/f45d6036e4210e1fbe52f6a582c540a6e1877056
@@ -0,0 +1,70 @@
+# RC-042: Review of BC-001
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-042 |
+| CrossReference | [BC-001], [QC-BC-001], [QC-LANG-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [BC-001]
+- Checklist used: [QC-BC-001], and [QC-LANG-001] (language and domain)
+- Scope: full review
+- Language and domain: en / it
+- Language reviewer: none (the stand-in for S01 reads the language and knows the domain)
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Checklist Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | ROI/Cost-Benefit analysis is quantitative, or where qualitative, is explicitly justified | Pass | The Cost-Benefit table is qualitative and says why: no revenue is expected, so a quantitative return does not apply. |
+| 2 | Risks are identified with documented impact and mitigation | Pass | The Risks table has 5 risks, each with an impact and a mitigation. |
+| 3 | Success criteria are measurable, stating explicit targets rather than vague aspirations | Pass | The Success Criteria table has 8 rows with numeric targets and a measure each (for example 0 third-party packages; 5 commands or fewer). |
+| 4 | Scope explicitly separates In Scope vs Out of Scope | Pass | Scope has `### In Scope` and `### Out of Scope`; the executable is added to In Scope and the installer, signing and release upload are named in Out of Scope. |
+| 5 | Stakeholders are cross-referenced to Stakeholder Analysis IDs rather than re-described inline | Pass | The Stakeholders table cites S01, S02 and S03 and does not re-describe their roles. |
+| 6 | Methodology and quality-standard foundation are stated explicitly (e.g. ISO/IEC 25010, Larman) | Pass | Methodological and Standards Foundation names the framework, Larman, ISO/IEC 25010:2023 and PEP 8. |
+| 7 | Assumptions and constraints are explicit and clearly distinguished from one another | Pass | Assumptions and Constraints are separate sections. |
+| 8 | Document supports executive decision-making with a clear, unambiguous recommendation | Pass | Recommendation reads `Proceed` with a one-sentence rationale. |
+
+## Language and Domain Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | The Metadata table has a `Language` row and a `Domain` row, and neither is a placeholder | Pass | Metadata of BC-001 has `Language | en` and `Domain | it`; neither is a placeholder. |
+| 2 | `Language` is a BCP 47 code and `Domain` is a value from the registry's domain list | Pass | `en` is a BCP 47 code and `it` is a value of the registry's domain list. |
+| 3 | The content (prose and table cells) is written in the stated language | Pass | Prose and table cells are English; Malay and Danish words and file names appear only as quoted data. |
+| 4 | The register matches the one the registry gives for the artifact type | Pass | The registry gives `IT Executive English` for this type; the text is written for that reader. |
+| 5 | Domain terms are the PO terms of the domain's dictionary, with no synonyms | Pass | The PO terms of DICT-001 (flash card, word list, words to learn, study session) are used and no synonym replaces them. |
+| 6 | Metadata keys, section headings, IDs and statuses are in English | Pass | Metadata keys, headings, IDs and statuses are English. |
+| 7 | No translated twin (`<name>.<language>.md`) exists beside the document | Pass | No translated twin exists beside the document (checked in `docs/`). |
+| 8 | A change of language or domain since the previous accepted version has a Version History row and was reviewed again | Pass | First version; no change of language or domain since an earlier accepted version. |
+| 9 | A reviewer competent in the domain, and in the language, has confirmed that the domain terms are used correctly | Pass | The reviewer is the stand-in for S01, the Product Owner, who reads English and knows the IT domain; no second language reviewer is needed (review-checklist-process step 2). Open to challenge by S01 in the end-of-run summary. |
+| 10 | Abbreviations are spelled out on first use, in the stated language | Fail | CSV, MVVM, UI, PO and similar abbreviations are not all spelled out at first use; Optional, deferred. |
+
+## Overall Verdict
+
+Go — every Mandatory criterion passes; the failed Optional criteria are deferred with an action item. Reviewed by AI as stand-in for S01; not independent of the author.
+
+Full review (not a delta) after the change for MIL-009 (executable).
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| Criterion LANG 10 (Optional) failed: Abbreviations are spelled out on first use, in the stated language; fix or accept the deferral | S01 | 2026-10-21 |
+
+---
+
+[BC-001]: ../../../docs/business-case.md
+[QC-BC-001]: ../../../framework/qc/artifacts/qc-business-case.md
+[QC-LANG-001]: ../../../framework/qc/artifacts/qc-language-domain.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,70 @@
+# RC-043: Review of US-001
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-043 |
+| CrossReference | [US-001], [QC-US-001], [QC-LANG-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [US-001]
+- Checklist used: [QC-US-001], and [QC-LANG-001] (language and domain)
+- Scope: full review
+- Language and domain: en / it
+- Language reviewer: none (the stand-in for S01 reads the language and knows the domain)
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Checklist Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | Follows INVEST criteria (Independent, Negotiable, Valuable, Estimable, Small, Testable) | Pass | INVEST Check section confirms the six properties for all seven stories and states one dependency exception. |
+| 2 | Written in "As a / I want / So that" form | Pass | All seven stories use `As a / I want / so that`. |
+| 3 | Clear, testable acceptance criteria are included | Pass | Every story, including US-001.07, has Given/When/Then acceptance criteria. |
+| 4 | Traceable to a use case or epic | Pass | Every story traces to a milestone and to UC-001. |
+| 5 | Story is sized to fit within a single iteration | Pass | Each story is one milestone, which fits one iteration. |
+| 6 | Story statement avoids technical implementation detail | Fail | Acceptance criteria name `words_to_learn.csv`, `ms-da.csv` and 3000 ms, which are assignment-mandated names and values rather than pure behaviour. |
+| 7 | Role named in the story matches an actor defined in the Use Case Diagram | Pass | The role `Learner` is the actor of UCD-001. |
+
+## Language and Domain Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | The Metadata table has a `Language` row and a `Domain` row, and neither is a placeholder | Pass | Metadata of US-001 has `Language | en` and `Domain | it`; neither is a placeholder. |
+| 2 | `Language` is a BCP 47 code and `Domain` is a value from the registry's domain list | Pass | `en` is a BCP 47 code and `it` is a value of the registry's domain list. |
+| 3 | The content (prose and table cells) is written in the stated language | Pass | Prose and table cells are English; Malay and Danish words and file names appear only as quoted data. |
+| 4 | The register matches the one the registry gives for the artifact type | Pass | The registry gives `IT Professional English` for this type; the text is written for that reader. |
+| 5 | Domain terms are the PO terms of the domain's dictionary, with no synonyms | Pass | The PO terms of DICT-001 (flash card, word list, words to learn, study session) are used and no synonym replaces them. |
+| 6 | Metadata keys, section headings, IDs and statuses are in English | Pass | Metadata keys, headings, IDs and statuses are English. |
+| 7 | No translated twin (`<name>.<language>.md`) exists beside the document | Pass | No translated twin exists beside the document (checked in `docs/`). |
+| 8 | A change of language or domain since the previous accepted version has a Version History row and was reviewed again | Pass | First version; no change of language or domain since an earlier accepted version. |
+| 9 | A reviewer competent in the domain, and in the language, has confirmed that the domain terms are used correctly | Pass | The reviewer is the stand-in for S01, the Product Owner, who reads English and knows the IT domain; no second language reviewer is needed (review-checklist-process step 2). Open to challenge by S01 in the end-of-run summary. |
+| 10 | Abbreviations are spelled out on first use, in the stated language | Fail | CSV, MVVM, UI, PO and similar abbreviations are not all spelled out at first use; Optional, deferred. |
+
+## Overall Verdict
+
+Go — every Mandatory criterion passes; the failed Optional criteria are deferred with an action item. Reviewed by AI as stand-in for S01; not independent of the author.
+
+Full review (not a delta) after the change for MIL-009 (executable).
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| Criterion 6 (Optional) failed: Story statement avoids technical implementation detail; fix or accept the deferral | S01 | 2026-10-21 |
+| Criterion LANG 10 (Optional) failed: Abbreviations are spelled out on first use, in the stated language; fix or accept the deferral | S01 | 2026-10-21 |
+
+---
+
+[US-001]: ../../../docs/user-stories.md
+[QC-US-001]: ../../../framework/qc/artifacts/qc-user-story.md
+[QC-LANG-001]: ../../../framework/qc/artifacts/qc-language-domain.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,54 @@
+# RC-044: Review of PP-001
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-044 |
+| CrossReference | [PP-001], [QC-LANG-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [PP-001]
+- Checklist used: none exists for this type, and [QC-LANG-001] (language and domain)
+- Scope: full review
+- Language and domain: en / it
+- Language reviewer: none (the stand-in for S01 reads the language and knows the domain)
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Language and Domain Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | The Metadata table has a `Language` row and a `Domain` row, and neither is a placeholder | Pass | Metadata of PP-001 has `Language | en` and `Domain | it`; neither is a placeholder. |
+| 2 | `Language` is a BCP 47 code and `Domain` is a value from the registry's domain list | Pass | `en` is a BCP 47 code and `it` is a value of the registry's domain list. |
+| 3 | The content (prose and table cells) is written in the stated language | Pass | Prose and table cells are English; Malay and Danish words and file names appear only as quoted data. |
+| 4 | The register matches the one the registry gives for the artifact type | Pass | The registry gives `IT Executive English` for this type; the text is written for that reader. |
+| 5 | Domain terms are the PO terms of the domain's dictionary, with no synonyms | Pass | The PO terms of DICT-001 (flash card, word list, words to learn, study session) are used and no synonym replaces them. |
+| 6 | Metadata keys, section headings, IDs and statuses are in English | Pass | Metadata keys, headings, IDs and statuses are English. |
+| 7 | No translated twin (`<name>.<language>.md`) exists beside the document | Pass | No translated twin exists beside the document (checked in `docs/`). |
+| 8 | A change of language or domain since the previous accepted version has a Version History row and was reviewed again | Pass | First version; no change of language or domain since an earlier accepted version. |
+| 9 | A reviewer competent in the domain, and in the language, has confirmed that the domain terms are used correctly | Pass | The reviewer is the stand-in for S01, the Product Owner, who reads English and knows the IT domain; no second language reviewer is needed (review-checklist-process step 2). Open to challenge by S01 in the end-of-run summary. |
+| 10 | Abbreviations are spelled out on first use, in the stated language | Fail | CSV, MVVM, UI, PO and similar abbreviations are not all spelled out at first use; Optional, deferred. |
+
+## Overall Verdict
+
+Go — no checklist exists for the Project Plan type; the plan now schedules MIL-009 on 2026-10-16 inside the Business Case constraint (delivery by 2026-10-21) and the language checklist passes. Accepted as the Product Owner's stand-in, after the Product Owner accepted the MIL-009 plan in chat. Reviewed by AI as stand-in for S01; not independent of the author.
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| Criterion LANG 10 (Optional) failed: Abbreviations are spelled out on first use, in the stated language; fix or accept the deferral | S01 | 2026-10-21 |
+
+---
+
+[PP-001]: ../../../docs/project-plan.md
+[QC-LANG-001]: ../../../framework/qc/artifacts/qc-language-domain.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,55 @@
+# RC-045: Review of DCD-001
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-045 |
+| CrossReference | [DCD-001], [QC-DCD-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [DCD-001]
+- Checklist used: [QC-DCD-001]
+- Scope: full review
+- Language and domain: n/a (technical type)
+- Language reviewer: none
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Checklist Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | SOLID principles applied; no god classes with excessive responsibilities | Pass | Each class has one responsibility (Class Table); view, view model, session, deck and adapters are separate. |
+| 2 | Visibility markers correct and consistent (`+` public, `-` private, `#` protected) | Pass | Visibility markers are on all members. |
+| 3 | Relationships correctly distinguished: Association vs Aggregation vs Composition vs Dependency | Pass | Composition (Vocabulary and Card), aggregation (Deck), association, dependency and realization are distinguished. |
+| 4 | Multiplicities and navigability specified on all associations | Pass | Multiplicities are on the associations; dependencies are directed. |
+| 5 | Applied design patterns are annotated explicitly (e.g. Singleton, Factory, Strategy) | Pass | Pattern Annotations table lists Repository, Ports and adapters, Null Object, MVVM, Composition root. |
+| 6 | Method signatures are traceable to Operation Contracts and/or design Sequence Diagrams | Pass | Method Traceability table is unchanged; the packaging support (`runtime.py`, `tools/`) is stated in Purpose and Scope as having no system operation and not being a design class, so it needs no row. |
+| 7 | Class names and structure remain consistent with the Domain Model concepts they refine | Pass | The Class Table has a `Refines` column; names match the dictionary's IT terms; no class was added. |
+| 8 | No circular dependencies between classes/packages unless explicitly justified | Pass | Dependency Check states there are no cycles. |
+| 9 | Dependencies between classes and packages point inward: domain classes do not depend on infrastructure, delivery or framework classes | Pass | Domain imports nothing; infrastructure and presentation point inward (Dependency Check). |
+
+## Overall Verdict
+
+Go — every Mandatory criterion passes. Reviewed by AI as stand-in for S01; not independent of the author.
+
+Full review (not a delta) after the change for MIL-009 (executable).
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| none | - | - |
+
+---
+
+[DCD-001]: ../../../docs/dcd.md
+[QC-DCD-001]: ../../../framework/qc/artifacts/qc-dcd.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,55 @@
+# RC-046: Review of ADR-0004
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-046 |
+| CrossReference | [ADR-0004], [QC-ADR-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [ADR-0004]
+- Checklist used: [QC-ADR-001]
+- Scope: full review
+- Language and domain: n/a (technical type)
+- Language reviewer: none
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Checklist Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | The ID (`ADR-NNNN`, 4-digit) matches the filename number and is unique and sequential | Pass | ID `ADR-0004` matches the file name number. |
+| 2 | Context states the problem, the forces (cost, risk, constraints) and the options evaluated | Pass | Context states the need (US-001.07), the forces and three options. |
+| 3 | At least two options were considered, or the absence of alternatives is justified | Pass | Three options were considered. |
+| 4 | Decision is stated in one or two unhedged sentences and matches one of the evaluated options | Pass | Decision is stated plainly and is option 2. |
+| 5 | Consequences list both Positive and Negative outcomes | Pass | Consequences have Positive and Negative lists. |
+| 6 | Quality-attribute impacts (performance, security, portability, etc.) are named where the decision affects them | Pass | Portability (Windows only, cross-compile limit), start-up time and size are named. |
+| 7 | Affected Artifacts are listed as links, and each is also in CrossReference | Pass | Affected Artifacts BC-001, US-001 and DCD-001 are listed and in CrossReference. |
+| 8 | Status changes are recorded as new Version History rows with a Change summary and Commit link; retained rows are never rewritten | Pass | First version; no status change yet. |
+| 9 | A superseded or deprecated ADR names its successor, and the successor links back | N-A | Not superseded or deprecated. |
+| 10 | Written in business or technical language of the domain, with no unexplained abbreviations | Pass | Written in plain technical English; `sys.frozen`, `sys.executable` are code names. |
+| 11 | Status is one of `Proposed`, `Accepted`, `Rejected`, `Deprecated` or `Superseded by ADR-NNNN`, and every Version History row uses only these | Pass | Status `Proposed` before this review; uses an allowed value. |
+
+## Overall Verdict
+
+Go — every Mandatory criterion passes. Reviewed by AI as stand-in for S01; not independent of the author.
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| none | - | - |
+
+---
+
+[ADR-0004]: ../../../docs/adr/adr-0004-pyinstaller-executable.md
+[QC-ADR-001]: ../../../framework/qc/artifacts/qc-adr.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -0,0 +1,66 @@
+# RC-047: Review of MIL-009
+
+## Metadata
+| Key | Value |
+| --- | --- |
+| ID | RC-047 |
+| CrossReference | [MIL-009], [QC-MIL-001], [QC-LANG-001] |
+
+## Version History
+| Date | Status | Author | Reviewer | Change | Commit |
+| --- | --- | --- | --- | --- | --- |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [768ca34] |
+
+---
+
+## Artifact Under Review
+
+- Instance reviewed: [MIL-009]
+- Checklist used: [QC-MIL-001], and [QC-LANG-001] (language and domain)
+- Scope: full review
+- Language and domain: en / it
+- Language reviewer: none (the stand-in for S01 reads the language and knows the domain)
+- Mode: autonomous, not independent (reviewer is `AI as stand-in for S01`)
+
+## Checklist Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | A concrete deliverable is defined for every gate | Pass | A concrete Deliverable section names the output evaluated at the gate. |
+| 2 | Explicit Go/No-Go criteria are stated for each gate | Pass | The Go / No-Go table has numbered criteria, each with an observable Go and No-Go. |
+| 3 | Dependencies on other milestones are explicitly mapped | Pass | Dependencies table names the milestone or documents it needs. |
+| 4 | Each milestone is traceable to a Business Case objective or KPI | Pass | Traceability table maps to Business Case objectives and the user story. |
+| 5 | Milestone owner and approving reviewer are identified | Pass | Ownership table names S01 as owner and approving reviewer. |
+| 6 | Milestone has a defined target date consistent with project constraints | Pass | Target Date lies inside the 2026-10-14 to 2026-10-21 window of the Business Case. |
+
+## Language and Domain Results
+
+| # | Criterion | Status | Evidence/Notes |
+| --- | --- | --- | --- |
+| 1 | The Metadata table has a `Language` row and a `Domain` row, and neither is a placeholder | Pass | Metadata of MIL-009 has `Language | en` and `Domain | it`; neither is a placeholder. |
+| 2 | `Language` is a BCP 47 code and `Domain` is a value from the registry's domain list | Pass | `en` is a BCP 47 code and `it` is a value of the registry's domain list. |
+| 3 | The content (prose and table cells) is written in the stated language | Pass | Prose and table cells are English; Malay and Danish words and file names appear only as quoted data. |
+| 4 | The register matches the one the registry gives for the artifact type | Pass | The registry gives `IT Executive English` for this type; the text is written for that reader. |
+| 5 | Domain terms are the PO terms of the domain's dictionary, with no synonyms | Pass | The PO terms of DICT-001 (flash card, word list, words to learn, study session) are used and no synonym replaces them. |
+| 6 | Metadata keys, section headings, IDs and statuses are in English | Pass | Metadata keys, headings, IDs and statuses are English. |
+| 7 | No translated twin (`<name>.<language>.md`) exists beside the document | Pass | No translated twin exists beside the document (checked in `docs/`). |
+| 8 | A change of language or domain since the previous accepted version has a Version History row and was reviewed again | Pass | First version; no change of language or domain since an earlier accepted version. |
+| 9 | A reviewer competent in the domain, and in the language, has confirmed that the domain terms are used correctly | Pass | The reviewer is the stand-in for S01, the Product Owner, who reads English and knows the IT domain; no second language reviewer is needed (review-checklist-process step 2). Open to challenge by S01 in the end-of-run summary. |
+| 10 | Abbreviations are spelled out on first use, in the stated language | Fail | CSV, MVVM, UI, PO and similar abbreviations are not all spelled out at first use; Optional, deferred. |
+
+## Overall Verdict
+
+Go — every Mandatory criterion passes; the failed Optional criteria are deferred with an action item. Reviewed by AI as stand-in for S01; not independent of the author.
+
+## Action Items
+
+| Action | Owner | Due |
+| --- | --- | --- |
+| Criterion LANG 10 (Optional) failed: Abbreviations are spelled out on first use, in the stated language; fix or accept the deferral | S01 | 2026-10-21 |
+
+---
+
+[MIL-009]: ../../../docs/milestones/mil-009-executable.md
+[QC-MIL-001]: ../../../framework/qc/artifacts/qc-milestones-gateways.md
+[QC-LANG-001]: ../../../framework/qc/artifacts/qc-language-domain.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -10,7 +10,8 @@
 | Date | Status | Author | Reviewer | Change | Commit |
 | --- | --- | --- | --- | --- | --- |
 | 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [c3fbcbc] |
-| 2026-10-14 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added MIL-008 and the review records of the level choice (no checklist exists for this type) | [0b983b7] |
+| 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added MIL-008 and the review records of the level choice (no checklist exists for this type) | [0b983b7] |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added the rows and review records of the executable request (no checklist exists for this type) | [768ca34] |

 ---

@@ -24,17 +25,17 @@ updated whenever an artifact instance is created or reviewed.

 | Artifact Instance | Type | Language | Domain | Upstream (Backward Link) | Downstream (Forward Link) | Last Reviewed (RC-ID) |
 | --- | --- | --- | --- | --- | --- | --- |
-| BC-001 | Business Case | en | it | - | SA-001, PP-001 | [RC-030] |
+| BC-001 | Business Case | en | it | - | SA-001, PP-001 | [RC-042] |
 | SA-001 | Stakeholder Analysis | en | it | BC-001 | PP-001, DICT-001 | [RC-002] |
-| PP-001 | Project Plan | en | it | BC-001, SA-001 | MIL-001 to MIL-008 | [RC-032] |
+| PP-001 | Project Plan | en | it | BC-001, SA-001 | MIL-001 to MIL-008 | [RC-044] |
 | UCD-001 | Use Case Diagram | en | it | SA-001, BC-001 | US-001, UC-001 | [RC-004] |
-| US-001 | User Story | en | it | UCD-001, BC-001 | MIL-002 to MIL-006, MIL-008 | [RC-031] |
+| US-001 | User Story | en | it | UCD-001, BC-001 | MIL-002 to MIL-006, MIL-008 | [RC-043] |
 | UC-001 | Use Case | en | it | UCD-001, US-001 | DM-001, SSD-001 | [RC-033] |
 | DM-001 | Domain Model | en | it | UC-001 | SSD-001, DICT-001, DCD-001 | [RC-034] |
 | SSD-001 | System Sequence Diagram | en | it | UC-001, DM-001 | DCD-001 | [RC-036] |
 | OC-001 | Operation Contract | - | - | SSD-001, DM-001 | SD-001 | [RC-037] |
 | SD-001 | Sequence Diagram | - | - | OC-001 | DCD-001 | [RC-038] |
-| DCD-001 | Design Class Diagram | - | - | DM-001, SD-001, DICT-001 | src/flash_card | [RC-039] |
+| DCD-001 | Design Class Diagram | - | - | DM-001, SD-001, DICT-001 | src/flash_card | [RC-045] |
 | DICT-001 | Domain Dictionary | en | it | BC-001, SA-001, DM-001 | DCD-001 | [RC-035] |
 | ADR-0001 | Architecture Decision Record | - | - | BC-001 | MIL-001, MIL-003, MIL-006 | [RC-013] |
 | ADR-0002 | Architecture Decision Record | - | - | BC-001 | DCD-001 | [RC-014] |
@@ -47,6 +48,8 @@ updated whenever an artifact instance is created or reviewed.
 | MIL-006 | Milestone / Gateway | en | it | BC-001, PP-001 | tasks of MIL-006 | [RC-021] |
 | MIL-007 | Milestone / Gateway | en | it | BC-001, PP-001 | tasks of MIL-007 | [RC-022] |
 | MIL-008 | Milestone / Gateway | en | it | BC-001, PP-001 | tasks of MIL-008 | [RC-040] |
+| MIL-009 | Milestone / Gateway | en | it | BC-001 | - | [RC-047] |
+| ADR-0004 | Architecture Decision Record | - | - | BC-001 | - | [RC-046] |
 | PY-MIL-001 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-023] |
 | PY-MIL-002 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-024] |
 | PY-MIL-003 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-025] |
@@ -54,6 +57,7 @@ updated whenever an artifact instance is created or reviewed.
 | PY-MIL-005 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-027] |
 | PY-MIL-006 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-028] |
 | PY-MIL-007 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-029] |
+| PY-MIL-008 | Python Source Code | - | - | DCD-001, ADR-0002 | - | [RC-041] |

 ## Coverage Notes

@@ -135,3 +139,14 @@ updated whenever an artifact instance is created or reviewed.
 [RC-040]: ./reviews/rc-040-mil-008.md
 [MIL-008]: ../milestones/mil-008-choose-level.md
 [0b983b7]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/0b983b7fdadf6958717c1d0227d0f2ebe037f7b1
+[RC-041]: ./reviews/rc-041-py-mil-008.md
+[PY-MIL-008]: ../../src/flash_card
+[RC-042]: ./reviews/rc-042-business-case-exe.md
+[RC-043]: ./reviews/rc-043-user-stories-exe.md
+[RC-044]: ./reviews/rc-044-project-plan-exe.md
+[RC-045]: ./reviews/rc-045-design-class-diagram-exe.md
+[RC-046]: ./reviews/rc-046-adr-0004.md
+[RC-047]: ./reviews/rc-047-mil-009.md
+[MIL-009]: ../milestones/mil-009-executable.md
+[ADR-0004]: ../adr/adr-0004-pyinstaller-executable.md
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -4,7 +4,7 @@
 | Key | Value |
 | --- | --- |
 | ID | US-001 |
-| CrossReference | [UCD-001], [BC-001], [UC-001], [MIL-002], [MIL-003], [MIL-004], [MIL-005], [MIL-006], [MIL-008] |
+| CrossReference | [UCD-001], [BC-001], [UC-001], [MIL-002], [MIL-003], [MIL-004], [MIL-005], [MIL-006], [MIL-008], [MIL-009] |
 | Language | en |
 | Domain | it |

@@ -12,13 +12,14 @@
 | Date | Status | Author | Reviewer | Change | Commit |
 | --- | --- | --- | --- | --- | --- |
 | 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Initial version | [c3fbcbc] |
-| 2026-10-14 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added US-001.06 Choose my level | [0b983b7] |
+| 2026-10-14 | Deprecated | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added US-001.06 Choose my level | [0b983b7] |
+| 2026-10-15 | Accepted | Jens Tirsvad Nielsen | AI as stand-in for S01 | Added US-001.07 Start without Python | [768ca34] |

 ---

 ## Purpose and Scope

-These stories cover the epic "Study Flash Cards": the four steps of the course assignment (the window, new cards, flipping the cards, saving progress), the spoken Malay word and the choice of level. Each story is delivered by one milestone. The wireframe the stories follow is in `docs/uc-001/wireframe.md`.
+These stories cover the epic "Study Flash Cards": the four steps of the course assignment (the window, new cards, flipping the cards, saving progress), the spoken Malay word the choice of level and the start without an installed Python. Each story is delivered by one milestone. The wireframe the stories follow is in `docs/uc-001/wireframe.md`.

 ## Story List

@@ -105,9 +106,24 @@ These stories cover the epic "Study Flash Cards": the four steps of the course a
 | --- | --- | --- |
 | [MIL-008], [UC-001] | fits one iteration | none |

+### US-001.07 — Start without Python
+
+**As a** Learner, **I want** to start the program by opening one executable file, **so that** I can learn without installing Python or a virtual environment.
+
+**Acceptance Criteria**
+
+- Given the executable was built, when I open it on a Windows computer without Python, then the level screen appears and no console window opens.
+- Given I mark a word as known in the executable, when I open it again, then the known word does not come up (my progress is kept in `data/words_to_learn.csv` next to the executable).
+- Given a `data/ms-da.csv` exists next to the executable, when the program starts, then that word list is used; otherwise the word list inside the executable is used.
+- Given the speech extra was included in the build and a network is available, when a card is shown, then the Malay word is spoken; without a network the program works silently.
+
+| Traces to | Size | INVEST exceptions |
+| --- | --- | --- |
+| [MIL-009], [UC-001] | fits one iteration | none |
+
 ## INVEST Check

-All six stories are Independent in delivery order (each builds on the one before but can be accepted alone), Negotiable in wording, Valuable to the Learner, Estimable, Small (one milestone each) and Testable (every criterion is Given/When/Then). Exception: US-001.02 to US-001.06 depend on the window of US-001.01, which is stated in their milestones' Dependencies.
+All seven stories are Independent in delivery order (each builds on the one before but can be accepted alone), Negotiable in wording, Valuable to the Learner, Estimable, Small (one milestone each) and Testable (every criterion is Given/When/Then). Exception: US-001.02 to US-001.07 depend on the window of US-001.01, which is stated in their milestones' Dependencies.

 ---

@@ -120,5 +136,7 @@ All six stories are Independent in delivery order (each builds on the one before
 [MIL-005]: ./milestones/mil-005-step4-save-progress.md
 [MIL-006]: ./milestones/mil-006-malay-speech.md
 [MIL-008]: ./milestones/mil-008-choose-level.md
+[MIL-009]: ./milestones/mil-009-executable.md
 [c3fbcbc]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/c3fbcbcfaf799e0ae34aed7e14d3bc8a62f03db3
 [0b983b7]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/0b983b7fdadf6958717c1d0227d0f2ebe037f7b1
+[768ca34]: https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/commit/768ca344c95d4f21e565e37ac96661c2f8dc8f54
@@ -6,7 +6,7 @@ Programmet er bygget i fire lag under `src/flash_card/`. Importer peger indad; `

 | Lag | Indeholder | Må importere |
 | --- | --- | --- |
-| `domain` | `Card`, `LanguagePair`, `Vocabulary`, `Deck`, fejlene og repository-grænsefladerne | ingenting |
+| `domain` | `Card`, `LanguagePair`, `Level`, `Vocabulary`, `Deck`, fejlene og repository-grænsefladerne | ingenting |
 | `application` | `StudySession`, `open_study_session`, `CardDto` og `SpeechPort` | `domain` |
 | `infrastructure` | CSV-repositories og tale-adaptere | `application`, `domain` |
 | `presentation` | `FlashCardView`, `FlashCardViewModel` (MVVM) og kompositionsroden `app.py` | alle lag |
@@ -21,6 +21,7 @@ Alle konstanter står i `src/flash_card/constants.py`.
 | Trin 2: læs CSV, tilfældigt ord ved hvert tryk | `CsvVocabularyRepository` og `read_word_list` (erstatter `pandas`), `Deck.draw`, `StudySession.next_card`, `FlashCardView.next_card` |
 | Trin 3: vend efter 3 sekunder | `FlashCardView.flip_card`, `FlashCardView.flip_timer` (`after`, `after_cancel`), `FlashCardViewModel.flip_card` |
 | Trin 4: kendte ord, `words_to_learn.csv` | `FlashCardView.is_known`, `StudySession.mark_known`, `CsvProgressRepository`, `open_study_session` |
+| Ekstra: niveauvalg | `Level`, `Deck.within`, `StudySession.next_card`, `open_study_session`, `LevelView`, `LevelViewModel`, `Application.start` i `presentation/app.py` |
 | Ekstra: tale | `SpeechPort`, `GttsSpeech`, `NullSpeech`, `build_speech` i `presentation/app.py` |

 ## Hvorfor en lagdelt løsning til en lille opgave?
@@ -6,7 +6,7 @@ Program dibina dalam empat lapisan di bawah `src/flash_card/`. Import menuju ke

 | Lapisan | Mengandungi | Boleh mengimport |
 | --- | --- | --- |
-| `domain` | `Card`, `LanguagePair`, `Vocabulary`, `Deck`, ralat dan antara muka repositori | tiada |
+| `domain` | `Card`, `LanguagePair`, `Level`, `Vocabulary`, `Deck`, ralat dan antara muka repositori | tiada |
 | `application` | `StudySession`, `open_study_session`, `CardDto` dan `SpeechPort` | `domain` |
 | `infrastructure` | Repositori CSV dan penyesuai suara | `application`, `domain` |
 | `presentation` | `FlashCardView`, `FlashCardViewModel` (MVVM) dan akar komposisi `app.py` | semua lapisan |
@@ -21,6 +21,7 @@ Semua pemalar terdapat dalam `src/flash_card/constants.py`.
 | Langkah 2: baca CSV, perkataan rawak setiap tekanan | `CsvVocabularyRepository` dan `read_word_list` (menggantikan `pandas`), `Deck.draw`, `StudySession.next_card`, `FlashCardView.next_card` |
 | Langkah 3: terbalik selepas 3 saat | `FlashCardView.flip_card`, `FlashCardView.flip_timer` (`after`, `after_cancel`), `FlashCardViewModel.flip_card` |
 | Langkah 4: perkataan diketahui, `words_to_learn.csv` | `FlashCardView.is_known`, `StudySession.mark_known`, `CsvProgressRepository`, `open_study_session` |
+| Tambahan: pilihan tahap | `Level`, `Deck.within`, `StudySession.next_card`, `open_study_session`, `LevelView`, `LevelViewModel`, `Application.start` dalam `presentation/app.py` |
 | Tambahan: suara | `SpeechPort`, `GttsSpeech`, `NullSpeech`, `build_speech` dalam `presentation/app.py` |

 ## Mengapa program berlapis untuk tugasan kecil?
@@ -6,7 +6,7 @@ The program is built in four layers under `src/flash_card/`. Imports point inwar

 | Layer | Holds | May import |
 | --- | --- | --- |
-| `domain` | `Card`, `LanguagePair`, `Vocabulary`, `Deck`, the errors and the repository interfaces | nothing |
+| `domain` | `Card`, `LanguagePair`, `Level`, `Vocabulary`, `Deck`, the errors and the repository interfaces | nothing |
 | `application` | `StudySession`, `open_study_session`, `CardDto` and `SpeechPort` | `domain` |
 | `infrastructure` | The CSV repositories and the speech adapters | `application`, `domain` |
 | `presentation` | `FlashCardView`, `FlashCardViewModel` (MVVM) and the composition root `app.py` | all layers |
@@ -21,6 +21,7 @@ All constants are in `src/flash_card/constants.py`.
 | Step 2: read the CSV, random word on each press | `CsvVocabularyRepository` and `read_word_list` (replace `pandas`), `Deck.draw`, `StudySession.next_card`, `FlashCardView.next_card` |
 | Step 3: flip after 3 seconds | `FlashCardView.flip_card`, `FlashCardView.flip_timer` (`after`, `after_cancel`), `FlashCardViewModel.flip_card` |
 | Step 4: known words, `words_to_learn.csv` | `FlashCardView.is_known`, `StudySession.mark_known`, `CsvProgressRepository`, `open_study_session` |
+| Extra: level choice | `Level`, `Deck.within`, `StudySession.next_card`, `open_study_session`, `LevelView`, `LevelViewModel`, `Application.start` in `presentation/app.py` |
 | Extra: speech | `SpeechPort`, `GttsSpeech`, `NullSpeech`, `build_speech` in `presentation/app.py` |

 ## Why a layered program for a small assignment?
@@ -4,6 +4,7 @@

 Et flashkortprogram til at lære de mest almindelige malajiske ord med danske oversættelser. Det er løsningen på dag 31 i Udemys kursus *100 Days of Code: The Complete Python Pro Bootcamp*.

+![Niveauskærmen: vælg 100 ord, 1000 ord eller alle ord](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-level.png)
 ![Forsiden af et kort: det malajiske ord](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-front.png)
 ![Bagsiden af kortet efter tre sekunder: den danske oversættelse](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-back.png)

@@ -22,6 +23,7 @@ Et flashkortprogram til at lære de mest almindelige malajiske ord med danske ov
 | 3. Vend kortene | Efter 3 sekunder viser kortet den danske oversættelse | MIL-004 |
 | 4. Gem din fremgang | Kendte ord fjernes, og resten gemmes i `words_to_learn.csv` | MIL-005 |
 | Ekstra | Det malajiske ord bliver udtalt (valgfrit) | MIL-006 |
+| Ekstra | Du vælger først et niveau: de første 100 ord, de første 1000 ord eller alle ord | MIL-008 |

 ## Til kursister

@@ -4,6 +4,7 @@

 Program kad imbas untuk mempelajari perkataan Melayu yang paling kerap digunakan beserta terjemahan Denmark. Ia ialah penyelesaian hari ke-31 kursus Udemy *100 Days of Code: The Complete Python Pro Bootcamp*.

+![Skrin tahap: pilih 100 perkataan, 1000 perkataan atau semua perkataan](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-level.png)
 ![Bahagian depan kad: perkataan Melayu](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-front.png)
 ![Bahagian belakang kad selepas tiga saat: terjemahan Denmark](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-back.png)

@@ -22,6 +23,7 @@ Program kad imbas untuk mempelajari perkataan Melayu yang paling kerap digunakan
 | 3. Terbalikkan kad | Selepas 3 saat kad menunjukkan terjemahan Denmark | MIL-004 |
 | 4. Simpan kemajuan anda | Perkataan yang diketahui dibuang dan selebihnya disimpan dalam `words_to_learn.csv` | MIL-005 |
 | Tambahan | Perkataan Melayu disebut (pilihan) | MIL-006 |
+| Tambahan | Anda memilih tahap dahulu: 100 perkataan pertama, 1000 perkataan pertama atau semua perkataan | MIL-008 |

 ## Untuk peserta kursus

@@ -4,6 +4,7 @@

 A flash card program for learning the most frequent Malay words with their Danish translations. It is the solution of day 31 of Udemy's *100 Days of Code: The Complete Python Pro Bootcamp*.

+![The level screen: choose 100 words, 1000 words or all words](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-level.png)
 ![The front of a card: the Malay word](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-front.png)
 ![The back of the card after three seconds: the Danish translation](https://git.tirsystem.com/Tirsvad-Udemy-100-days-of-code/031-flash-card/raw/branch/main/docs/images/screenshot-back.png)

@@ -22,6 +23,7 @@ A flash card program for learning the most frequent Malay words with their Danis
 | 3. Flip the cards | After 3 seconds the card shows the Danish translation | MIL-004 |
 | 4. Save your progress | Known words are removed and the rest is saved in `words_to_learn.csv` | MIL-005 |
 | Extra | The Malay word is spoken (optional) | MIL-006 |
+| Extra | You choose a level first: the first 100 words, the first 1000 words or all words | MIL-008 |

 ## For course participants

@@ -5,12 +5,13 @@
 ## Øv kortene

 1. Start programmet med `python -m flash_card`.
-2. Et kort viser et malajisk ord. Prøv at huske den danske oversættelse.
-3. Efter tre sekunder vender kortet og viser oversættelsen.
-4. Tryk på det grønne flueben, hvis du kendte ordet, eller på det røde kryds, hvis du ikke gjorde.
-5. Et nyt tilfældigt ord vises. Gentag.
+2. Vælg et niveau: **100 words**, **1000 words** eller **All words**. Ordlisten er sorteret fra det hyppigste til det sjældneste ord, så niveau 100 er de 100 hyppigste ord.
+3. Et kort viser et malajisk ord. Prøv at huske den danske oversættelse.
+4. Efter tre sekunder vender kortet og viser oversættelsen.
+5. Tryk på det grønne flueben, hvis du kendte ordet, eller på det røde kryds, hvis du ikke gjorde.
+6. Et nyt tilfældigt ord fra niveauet vises. Gentag.

-Et ord, du har markeret som kendt, fjernes og kommer ikke igen. Når der ikke er flere ord, står der "Nothing left to learn" på kortet.
+Et ord, du har markeret som kendt, fjernes og kommer ikke igen. Når der ikke er flere ord på niveauet, står der "Nothing left to learn" på kortet; start programmet igen og vælg et højere niveau.

 ## Din fremgang

@@ -18,6 +19,8 @@ Et ord, du har markeret som kendt, fjernes og kommer ikke igen. Når der ikke er
 - Næste start bruger den fil. Findes den ikke, bruger programmet hele ordlisten `data/ms-da.csv`.
 - For at starte forfra skal du slette `data/words_to_learn.csv`.
 - Det røde kryds ændrer aldrig filen.
+- Kendte ord forbliver kendte på alle niveauer. Niveauet begrænser kun, hvilke af de ord, du endnu ikke kender, der kommer frem, og filen beholder altid alle ord, du ikke kender, også dem uden for niveauet.
+- Denne liste har 1000 ord, så "1000 words" og "All words" giver de samme kort.

 ## Filformat

@@ -34,6 +37,7 @@ yang;hvilke
 | Indstilling | Betydning |
 | --- | --- |
 | `--data-dir <mappe>` | Mappen med `ms-da.csv` og `words_to_learn.csv` (standard: `data/` i repositoryet) |
+| `--level {100,1000,all}` | Start på det niveau og spring niveauskærmen over |
 | `--no-speech` | Udtal ikke det malajiske ord, selv om tilvalget til tale er installeret |

 ## Brug et andet sprogpar
@@ -5,12 +5,13 @@
 ## Berlatih dengan kad

 1. Mulakan program dengan `python -m flash_card`.
-2. Kad menunjukkan satu perkataan Melayu. Cuba ingat terjemahan Denmarknya.
-3. Selepas tiga saat kad terbalik dan menunjukkan terjemahan.
-4. Tekan tanda semak hijau jika anda tahu perkataan itu, atau pangkah merah jika tidak.
-5. Satu perkataan rawak baharu muncul. Ulang.
+2. Pilih tahap: **100 words**, **1000 words** atau **All words**. Senarai perkataan disusun daripada perkataan paling kerap kepada paling jarang, jadi tahap 100 ialah 100 perkataan paling kerap.
+3. Kad menunjukkan satu perkataan Melayu. Cuba ingat terjemahan Denmarknya.
+4. Selepas tiga saat kad terbalik dan menunjukkan terjemahan.
+5. Tekan tanda semak hijau jika anda tahu perkataan itu, atau pangkah merah jika tidak.
+6. Satu perkataan rawak baharu daripada tahap itu muncul. Ulang.

-Perkataan yang anda tandakan sebagai diketahui dibuang dan tidak muncul lagi. Apabila tiada perkataan lagi, kad menunjukkan "Nothing left to learn".
+Perkataan yang anda tandakan sebagai diketahui dibuang dan tidak muncul lagi. Apabila tiada perkataan lagi pada tahap itu, kad menunjukkan "Nothing left to learn"; mulakan program semula dan pilih tahap yang lebih tinggi.

 ## Kemajuan anda

@@ -18,6 +19,8 @@ Perkataan yang anda tandakan sebagai diketahui dibuang dan tidak muncul lagi. Ap
 - Permulaan seterusnya menggunakan fail itu. Jika ia tiada, program menggunakan senarai penuh `data/ms-da.csv`.
 - Untuk bermula semula, padam `data/words_to_learn.csv`.
 - Pangkah merah tidak pernah mengubah fail.
+- Perkataan yang diketahui kekal diketahui pada semua tahap. Tahap hanya mengehadkan perkataan yang belum anda ketahui yang muncul, dan fail sentiasa menyimpan semua perkataan yang anda belum tahu, termasuk yang di luar tahap.
+- Senarai ini mempunyai 1000 perkataan, jadi "1000 words" dan "All words" memberikan kad yang sama.

 ## Format fail

@@ -34,6 +37,7 @@ yang;hvilke
 | Pilihan | Maksud |
 | --- | --- |
 | `--data-dir <folder>` | Folder yang mengandungi `ms-da.csv` dan `words_to_learn.csv` (lalai: `data/` dalam repositori) |
+| `--level {100,1000,all}` | Mulakan pada tahap itu dan langkau skrin tahap |
 | `--no-speech` | Jangan sebut perkataan Melayu, walaupun tambahan suara dipasang |

 ## Gunakan pasangan bahasa lain
@@ -5,12 +5,13 @@
 ## Study the cards

 1. Start the program with `python -m flash_card`.
-2. A card shows a Malay word. Try to remember its Danish translation.
-3. After three seconds the card turns over and shows the translation.
-4. Press the green tick when you knew the word, or the red cross when you did not.
-5. A new random word appears. Repeat.
+2. Choose a level: **100 words**, **1000 words** or **All words**. The word list is ordered from the most to the least frequent word, so level 100 is the 100 most frequent words.
+3. A card shows a Malay word. Try to remember its Danish translation.
+4. After three seconds the card turns over and shows the translation.
+5. Press the green tick when you knew the word, or the red cross when you did not.
+6. A new random word of the level appears. Repeat.

-A word you marked as known is removed and does not come up again. When no word is left, the card says "Nothing left to learn".
+A word you marked as known is removed and does not come up again. When no word of the level is left, the card says "Nothing left to learn"; restart the program and choose a higher level.

 ## Your progress

@@ -18,6 +19,8 @@ A word you marked as known is removed and does not come up again. When no word i
 - The next start uses that file. If it does not exist, the program uses the full word list `data/ms-da.csv`.
 - To start again, delete `data/words_to_learn.csv`.
 - The red cross never changes the file.
+- Known words stay known at every level. The level only limits which of the words you do not know yet come up, and the file always keeps every word you do not know, also those outside the level.
+- This list has 1000 words, so "1000 words" and "All words" give the same cards.

 ## File format

@@ -34,6 +37,7 @@ yang;hvilke
 | Option | Meaning |
 | --- | --- |
 | `--data-dir <folder>` | Folder that holds `ms-da.csv` and `words_to_learn.csv` (default: `data/` of the repository) |
+| `--level {100,1000,all}` | Start at that level and skip the level screen |
 | `--no-speech` | Do not speak the Malay word, even when the speech extra is installed |

 ## Use another language pair
@@ -12,6 +12,7 @@ from flash_card.domain.entities.card import Card
 from flash_card.domain.entities.deck import Deck
 from flash_card.domain.repositories.progress_repository import ProgressRepository
 from flash_card.domain.repositories.vocabulary_repository import VocabularyRepository
+from flash_card.domain.value_objects.level import Level


 def open_study_session(
@@ -19,18 +20,24 @@ def open_study_session(
    progress_repository: ProgressRepository,
    speech: SpeechPort,
    choose: Callable[[Sequence[Card]], Card],
+    level: Level,
 ) -> StudySession:
    """Build a session over the saved words to learn, else the whole word list.

+    The word list is always loaded: its order defines which words a level holds.
+
    @param vocabulary_repository Source of the delivered word list.
    @param progress_repository Store of the words still to learn.
    @param speech Speaks the front text of each drawn card.
    @param choose Picks one card from a non-empty sequence.
+    @param level The level: the first words of the list that may be drawn.
    @return The new session.
-    @throws VocabularyError When the saved progress, or the word list when no
-        progress exists, cannot be loaded.
+    @throws VocabularyError When the word list or the saved progress cannot be
+        loaded.
    """
+    vocabulary = vocabulary_repository.load()
    deck = progress_repository.load()
    if deck is None:
-        deck = Deck.from_vocabulary(vocabulary_repository.load())
-    return StudySession(deck, choose, progress_repository, speech)
+        deck = Deck.from_vocabulary(vocabulary)
+    level_cards = level.first(vocabulary.cards)
+    return StudySession(deck, choose, progress_repository, speech, level_cards)
@@ -22,6 +22,7 @@ class StudySession:
        choose: Callable[[Sequence[Card]], Card],
        progress: ProgressRepository,
        speech: SpeechPort,
+        level_cards: Sequence[Card] | None = None,
    ) -> None:
        """Create a session.

@@ -29,7 +30,10 @@ class StudySession:
        @param choose Picks one card from a non-empty sequence.
        @param progress Store of the words still to learn.
        @param speech Speaks the front text of each drawn card.
+        @param level_cards The cards of the chosen level, or None for no limit.
        """
+        ## @brief The cards of the chosen level, or None for no limit.
+        self._level_cards = None if level_cards is None else tuple(level_cards)
        ## @brief Speaks the front text of each drawn card.
        self._speech = speech
        ## @brief Store of the words still to learn.
@@ -51,27 +55,36 @@ class StudySession:

    @property
    def remaining(self) -> int:
-        """Return the number of cards still to learn.
+        """Return the number of cards still to learn, of every level.

        @return The card count.
        """
        return len(self._deck)

+    @property
+    def remaining_in_level(self) -> int:
+        """Return the number of cards still to learn in the chosen level.
+
+        @return The card count.
+        """
+        return len(self._candidates())
+
    def next_card(self) -> CardDto | None:
        """Draw the next card, different from the current one when possible.

-        @return The new card, or None when the deck is empty.
+        @return The new card, or None when no card of the level is left.
        """
-        if self._deck.is_empty:
+        candidates = self._candidates()
+        if not candidates:
            self._current = None
            return None
-        card = self._deck.draw(self._choose, exclude=self._current)
+        card = self._deck.draw(self._choose, self._current, candidates)
        self._current = card
        self._speech.speak(card.front, self._deck.language_pair.front_code)
        return self._to_dto(card)

    def mark_known(self) -> None:
-        """Remove the current card from the deck and save the remaining cards.
+        """Remove the current card from the whole deck and save the whole deck.

        Does nothing and saves nothing when no card is shown.

@@ -83,6 +96,15 @@ class StudySession:
        self._progress.save(self._deck)
        self._current = None

+    def _candidates(self) -> tuple[Card, ...]:
+        """Return the cards that may be drawn.
+
+        @return The deck cards of the level, or all deck cards without a level.
+        """
+        if self._level_cards is None:
+            return self._deck.cards
+        return self._deck.within(self._level_cards)
+
    def _to_dto(self, card: Card | None) -> CardDto | None:
        """Map a card to its DTO.

@@ -20,6 +20,8 @@ WINDOW_PADDING: Final[int] = 50
 CARD_WIDTH: Final[int] = 800
 ## @brief Card height in pixels.
 CARD_HEIGHT: Final[int] = 526
+## @brief Level screen height in pixels: card plus the card window button row.
+LEVEL_SCREEN_HEIGHT: Final[int] = 526 + 100
 ## @brief Horizontal centre of the card in pixels.
 CARD_CENTER_X: Final[int] = 400
 ## @brief Vertical position of the language text in pixels.
@@ -43,6 +45,29 @@ LANGUAGE_NAMES: Final[Mapping[str, str]] = {"ms": "Malay", "da": "Danish"}
 ## @brief Message shown when no word is left to learn.
 FINISHED_MESSAGE: Final[str] = "Nothing left to learn"

+## @brief Title of the level screen.
+LEVEL_TITLE: Final[str] = "Choose your level"
+## @brief Level choices in display order: label and word limit (None = all words).
+LEVEL_CHOICES: Final[tuple[tuple[str, int | None], ...]] = (
+    ("100 words", 100),
+    ("1000 words", 1000),
+    ("All words", None),
+)
+## @brief Values of the --level option and the word limit each stands for.
+LEVEL_ARGUMENT_VALUES: Final[Mapping[str, int | None]] = {
+    "100": 100,
+    "1000": 1000,
+    "all": None,
+}
+## @brief Font of the level screen title: family, size, style.
+LEVEL_TITLE_FONT: Final[tuple[str, int, str]] = ("Arial", 40, "italic")
+## @brief Font of the level buttons: family, size, style.
+LEVEL_BUTTON_FONT: Final[tuple[str, int, str]] = ("Arial", 24, "bold")
+## @brief Width of every level button in characters.
+LEVEL_BUTTON_WIDTH: Final[int] = 14
+## @brief Background colour of the level buttons.
+LEVEL_BUTTON_COLOR: Final[str] = "white"
+
 ## @brief Delimiter of the CSV files.
 CSV_DELIMITER: Final[str] = ";"
 ## @brief Encoding of the CSV files; drops a leading byte order mark.
@@ -4,6 +4,7 @@
@brief The Deck entity: holds cards, draws one and removes one.
 """

+from collections import Counter
 from collections.abc import Callable, Iterable, Sequence

 from flash_card.domain.entities.card import Card
@@ -66,24 +67,43 @@ class Deck:
        """
        return len(self._cards)

+    def within(self, cards: Sequence[Card]) -> tuple[Card, ...]:
+        """Select the cards of the deck that also appear in the given cards.
+
+        A card counts as often as it appears in both.
+
+        @param cards The cards to match against, for example a level.
+        @return The matching cards in deck order.
+        """
+        allowed = Counter(cards)
+        selected: list[Card] = []
+        for card in self._cards:
+            if allowed[card] > 0:
+                allowed[card] -= 1
+                selected.append(card)
+        return tuple(selected)
+
    def draw(
        self,
        choose: Callable[[Sequence[Card]], Card],
        exclude: Card | None = None,
+        among: Sequence[Card] | None = None,
    ) -> Card:
        """Pick a card without removing it.

-        The pick avoids the excluded card when another card exists.
+        The pick avoids the excluded card when another candidate exists.

        @param choose Picks one card from a non-empty sequence.
        @param exclude A card to avoid, usually the one shown now.
+        @param among The candidates, or None for every card of the deck.
        @return The picked card.
-        @throws EmptyDeckError When the deck has no card.
+        @throws EmptyDeckError When there is no candidate.
        """
-        if not self._cards:
+        candidates = self._cards if among is None else list(among)
+        if not candidates:
            raise EmptyDeckError("cannot draw from an empty deck")
-        others = [card for card in self._cards if card != exclude]
-        return choose(others or self._cards)
+        others = [card for card in candidates if card != exclude]
+        return choose(others or candidates)

    def remove(self, card: Card) -> None:
        """Remove the first card equal to the given one.
@@ -0,0 +1,38 @@
+"""The level of a study session.
+
+@file level.py
+@brief The Level value object: how many of the first words a session uses.
+"""
+
+from collections.abc import Sequence
+from dataclasses import dataclass
+
+from flash_card.domain.entities.card import Card
+
+
+@dataclass(frozen=True, slots=True)
+class Level:
+    """A limit on the words of the list, counted from the most frequent."""
+
+    ## @brief Number of first words used, or None for all words.
+    limit: int | None
+
+    def __post_init__(self) -> None:
+        """Check the limit.
+
+        @return Nothing.
+        @throws ValueError When the limit is set and not above zero.
+        """
+        if self.limit is not None and self.limit <= 0:
+            raise ValueError(f"level limit must be above zero, got {self.limit}")
+
+    def first(self, cards: Sequence[Card]) -> tuple[Card, ...]:
+        """Select the cards of this level.
+
+        @param cards The word list in order from the most frequent word.
+        @return The first limit cards, or all cards when there is no limit or
+            the list is shorter.
+        """
+        if self.limit is None:
+            return tuple(cards)
+        return tuple(cards[: self.limit])
@@ -1,7 +1,7 @@
 """Entry point and composition root of the flash card program.

@file app.py
-@brief Creates the window, its view model and view, and runs the main loop.
+@brief Creates the window, shows the level screen and then the card screen.
 """

 import argparse
@@ -9,7 +9,7 @@ import importlib.util
 import logging
 import random
 import tkinter as tk
-from collections.abc import Sequence
+from collections.abc import Callable, Sequence
 from pathlib import Path
 from typing import NoReturn

@@ -17,10 +17,15 @@ from flash_card.application.ports.speech_port import SpeechPort
 from flash_card.application.use_cases.open_study_session import open_study_session
 from flash_card.constants import (
    DEFAULT_DATA_DIR,
+    LEVEL_ARGUMENT_VALUES,
    PROGRESS_FILE_NAME,
    VOCABULARY_FILE_NAME,
 )
+from flash_card.domain.entities.card import Card
 from flash_card.domain.errors.vocabulary_errors import VocabularyError
+from flash_card.domain.repositories.progress_repository import ProgressRepository
+from flash_card.domain.repositories.vocabulary_repository import VocabularyRepository
+from flash_card.domain.value_objects.level import Level
 from flash_card.infrastructure.persistence.csv_progress_repository import (
    CsvProgressRepository,
 )
@@ -33,6 +38,8 @@ from flash_card.presentation.features.flash_card.flash_card_view import FlashCar
 from flash_card.presentation.features.flash_card.flash_card_view_model import (
    FlashCardViewModel,
 )
+from flash_card.presentation.features.level.level_view import LevelView
+from flash_card.presentation.features.level.level_view_model import LevelViewModel

 ## @brief Logger of the entry point.
 _logger = logging.getLogger(__name__)
@@ -55,26 +62,141 @@ def build_speech(enabled: bool) -> SpeechPort:
    return NullSpeech()


+class Application:
+    """The running program: one window that shows the level screen, then the card."""
+
+    def __init__(
+        self,
+        root: tk.Tk,
+        vocabulary_repository: VocabularyRepository,
+        progress_repository: ProgressRepository,
+        speech: SpeechPort,
+        choose: Callable[[Sequence[Card]], Card] = random.choice,
+    ) -> None:
+        """Create the application on a window; no screen is shown yet.
+
+        @param root The Tk root window.
+        @param vocabulary_repository Source of the delivered word list.
+        @param progress_repository Store of the words still to learn.
+        @param speech Speaks the front text of each drawn card.
+        @param choose Picks one card from a non-empty sequence.
+        """
+        ## @brief The Tk root window.
+        self.root = root
+        ## @brief Source of the delivered word list.
+        self._vocabulary_repository = vocabulary_repository
+        ## @brief Store of the words still to learn.
+        self._progress_repository = progress_repository
+        ## @brief Speaks the front text of each drawn card.
+        self._speech = speech
+        ## @brief Picks one card from a non-empty sequence.
+        self._choose = choose
+        ## @brief The screen shown now, or None before the first one.
+        self._view: LevelView | FlashCardView | None = None
+        ## @brief The exit code of the program.
+        self._exit_code = 0
+
+    @property
+    def view(self) -> LevelView | FlashCardView | None:
+        """Return the screen shown now.
+
+        @return The level screen, the card screen, or None before the first.
+        """
+        return self._view
+
+    @property
+    def exit_code(self) -> int:
+        """Return the exit code the program ends with.
+
+        @return 0, or 1 when a session could not be started after a choice.
+        """
+        return self._exit_code
+
+    def show_level_screen(self) -> None:
+        """Show the level screen; a choice starts the session.
+
+        @return Nothing.
+        """
+        self._view = LevelView(self.root, LevelViewModel(), self._choose_level)
+
+    def start(self, level: Level) -> None:
+        """Open the session of a level and show the card screen on the window.
+
+        @param level The level the user chose.
+        @return Nothing.
+        @throws VocabularyError When the word list or saved progress is unusable.
+        """
+        session = open_study_session(
+            self._vocabulary_repository,
+            self._progress_repository,
+            self._speech,
+            self._choose,
+            level,
+        )
+        if isinstance(self._view, LevelView):
+            self._view.destroy()
+        self._view = FlashCardView(self.root, FlashCardViewModel(session))
+
+    def _choose_level(self, level: Level) -> None:
+        """Start the session after a click; close the window when it fails.
+
+        @param level The level the user chose.
+        @return Nothing.
+        """
+        try:
+            self.start(level)
+        except VocabularyError as error:
+            _logger.error("%s", error)
+            self._exit_code = 1
+            self.root.destroy()
+
+
 def create_application(
    data_dir: Path | None = None,
    speech_enabled: bool = True,
-) -> tuple[tk.Tk, FlashCardView]:
-    """Create the root window and the flash card view.
+    level: Level | None = None,
+) -> Application:
+    """Create the window and show the level screen, or start at the given level.

    @param data_dir Folder of the data files, or None for the default folder.
    @param speech_enabled False to run without spoken words.
-    @return The root window and the view built on it.
-    @throws VocabularyError When the word list or saved progress is unusable.
+    @param level The level to start at, or None to show the level screen.
+    @return The application, its window in the root attribute.
+    @throws VocabularyError When a level is given and the word list or saved
+        progress is unusable.
    """
    folder = DEFAULT_DATA_DIR if data_dir is None else data_dir
-    repository = CsvVocabularyRepository(folder / VOCABULARY_FILE_NAME)
-    progress = CsvProgressRepository(folder / PROGRESS_FILE_NAME)
-    speech = build_speech(speech_enabled)
-    session = open_study_session(repository, progress, speech, random.choice)
-    view_model = FlashCardViewModel(session)
    root = tk.Tk()
-    view = FlashCardView(root, view_model)
-    return root, view
+    application = Application(
+        root,
+        CsvVocabularyRepository(folder / VOCABULARY_FILE_NAME),
+        CsvProgressRepository(folder / PROGRESS_FILE_NAME),
+        build_speech(speech_enabled),
+    )
+    if level is None:
+        application.show_level_screen()
+        return application
+    try:
+        application.start(level)
+    except VocabularyError:
+        root.destroy()
+        raise
+    return application
+
+
+def _level_argument(value: str) -> Level:
+    """Convert a --level value to a level.
+
+    @param value One of the keys of LEVEL_ARGUMENT_VALUES.
+    @return The level.
+    @throws argparse.ArgumentTypeError When the value is unknown.
+    """
+    if value not in LEVEL_ARGUMENT_VALUES:
+        choices = ", ".join(LEVEL_ARGUMENT_VALUES)
+        raise argparse.ArgumentTypeError(
+            f"invalid level {value!r} (choose from {choices})"
+        )
+    return Level(LEVEL_ARGUMENT_VALUES[value])


 def main(argv: Sequence[str] | None = None) -> int:
@@ -95,14 +217,21 @@ def main(argv: Sequence[str] | None = None) -> int:
        action="store_true",
        help="do not speak the words",
    )
+    parser.add_argument(
+        "--level",
+        type=_level_argument,
+        default=None,
+        metavar="{" + ",".join(LEVEL_ARGUMENT_VALUES) + "}",
+        help="start at a level (100, 1000 or all words) and skip the level screen",
+    )
    args = parser.parse_args(argv)
    try:
-        root, _view = create_application(args.data_dir, not args.no_speech)
+        application = create_application(args.data_dir, not args.no_speech, args.level)
    except VocabularyError as error:
        _logger.error("%s", error)
        return 1
-    root.mainloop()
-    return 0
+    application.root.mainloop()
+    return application.exit_code


 def run() -> NoReturn:
@@ -20,8 +20,6 @@ from flash_card.constants import (
    LANGUAGE_FONT,
    LANGUAGE_TEXT_Y,
    RIGHT_IMAGE_NAME,
-    WINDOW_PADDING,
-    WINDOW_TITLE,
    WORD_FONT,
    WORD_TEXT_Y,
    WRONG_IMAGE_NAME,
@@ -30,6 +28,7 @@ from flash_card.presentation.features.flash_card.flash_card_view_model import (
    FlashCardViewModel,
    Side,
 )
+from flash_card.presentation.shared.window import configure_window


 def _load_image(root: tk.Tk, name: str) -> tk.PhotoImage:
@@ -60,8 +59,7 @@ class FlashCardView:
        self._root = root
        ## @brief The data source of the texts.
        self._view_model = view_model
-        root.title(WINDOW_TITLE)
-        root.config(padx=WINDOW_PADDING, pady=WINDOW_PADDING, bg=BACKGROUND_COLOR)
+        configure_window(root)

        # Kept as attributes: Tkinter drops an image nothing refers to.
        ## @brief Card front image, kept so Tk does not drop it.
@@ -0,0 +1 @@
+"""The level feature: its view and view model."""
@@ -0,0 +1,97 @@
+"""View of the level screen.
+
+@file level_view.py
+@brief Builds the level screen of the wireframe: a title and one button per level.
+"""
+
+import tkinter as tk
+from collections.abc import Callable
+
+from flash_card.constants import (
+    BACKGROUND_COLOR,
+    CARD_WIDTH,
+    LEVEL_BUTTON_COLOR,
+    LEVEL_BUTTON_FONT,
+    LEVEL_BUTTON_WIDTH,
+    LEVEL_SCREEN_HEIGHT,
+    LEVEL_TITLE_FONT,
+)
+from flash_card.domain.value_objects.level import Level
+from flash_card.presentation.features.level.level_view_model import (
+    LevelChoice,
+    LevelViewModel,
+)
+
+
+class LevelView:
+    """The level screen: a title and a button for each level."""
+
+    def __init__(
+        self,
+        root: tk.Tk,
+        view_model: LevelViewModel,
+        on_choose: Callable[[Level], None],
+    ) -> None:
+        """Build the screen on a root.
+
+        @param root The Tk root window.
+        @param view_model The data to show.
+        @param on_choose Called with the level the user picked.
+        """
+        ## @brief Called with the level the user picked.
+        self._on_choose = on_choose
+        ## @brief Frame that holds the whole screen.
+        self.frame = tk.Frame(
+            root, bg=BACKGROUND_COLOR, width=CARD_WIDTH, height=LEVEL_SCREEN_HEIGHT
+        )
+        self.frame.grid(row=0, column=0)
+        self.frame.grid_propagate(False)
+        self.frame.grid_columnconfigure(0, weight=1)
+        for row in range(len(view_model.choices) + 1):
+            self.frame.grid_rowconfigure(row, weight=1)
+        ## @brief The title text.
+        self.title_label = tk.Label(
+            self.frame,
+            text=view_model.title,
+            font=LEVEL_TITLE_FONT,
+            bg=BACKGROUND_COLOR,
+        )
+        self.title_label.grid(row=0, column=0)
+        ## @brief One button per choice, in display order.
+        self.buttons = tuple(
+            self._build_button(row, choice)
+            for row, choice in enumerate(view_model.choices, start=1)
+        )
+
+    def choose(self, level: Level) -> None:
+        """Forward the chosen level.
+
+        @param level The level the user picked.
+        @return Nothing.
+        """
+        self._on_choose(level)
+
+    def destroy(self) -> None:
+        """Remove the screen from the window.
+
+        @return Nothing.
+        """
+        self.frame.destroy()
+
+    def _build_button(self, row: int, choice: LevelChoice) -> tk.Button:
+        """Create the button of one choice.
+
+        @param row Grid row of the button.
+        @param choice The choice the button stands for.
+        @return The button, placed in the frame.
+        """
+        button = tk.Button(
+            self.frame,
+            text=choice.label,
+            font=LEVEL_BUTTON_FONT,
+            width=LEVEL_BUTTON_WIDTH,
+            bg=LEVEL_BUTTON_COLOR,
+            command=lambda: self.choose(choice.level),
+        )
+        button.grid(row=row, column=0)
+        return button
@@ -0,0 +1,53 @@
+"""View model of the level screen.
+
+@file level_view_model.py
+@brief Exposes the title and the level choices; free of any Tkinter import.
+"""
+
+from collections.abc import Sequence
+from dataclasses import dataclass
+
+from flash_card.constants import LEVEL_CHOICES, LEVEL_TITLE
+from flash_card.domain.value_objects.level import Level
+
+
+@dataclass(frozen=True, slots=True)
+class LevelChoice:
+    """One choice on the level screen."""
+
+    ## @brief Text of the choice.
+    label: str
+    ## @brief The level the choice stands for.
+    level: Level
+
+
+class LevelViewModel:
+    """Data the level view shows."""
+
+    def __init__(self, choices: Sequence[tuple[str, int | None]] | None = None) -> None:
+        """Create the view model.
+
+        @param choices Label and word limit of each choice, or None for the
+            choices of the program.
+        """
+        source = LEVEL_CHOICES if choices is None else choices
+        ## @brief The choices in display order.
+        self._choices = tuple(
+            LevelChoice(label, Level(limit)) for label, limit in source
+        )
+
+    @property
+    def title(self) -> str:
+        """Return the title of the screen.
+
+        @return The title text.
+        """
+        return LEVEL_TITLE
+
+    @property
+    def choices(self) -> tuple[LevelChoice, ...]:
+        """Return the choices.
+
+        @return The choices in display order.
+        """
+        return self._choices
@@ -0,0 +1 @@
+"""Code shared by the screens of the presentation layer."""
@@ -0,0 +1,19 @@
+"""Window set-up shared by the screens.
+
+@file window.py
+@brief Gives the root window its title, padding and background.
+"""
+
+import tkinter as tk
+
+from flash_card.constants import BACKGROUND_COLOR, WINDOW_PADDING, WINDOW_TITLE
+
+
+def configure_window(root: tk.Tk) -> None:
+    """Set the title, padding and background of the window.
+
+    @param root The Tk root window.
+    @return Nothing.
+    """
+    root.title(WINDOW_TITLE)
+    root.config(padx=WINDOW_PADDING, pady=WINDOW_PADDING, bg=BACKGROUND_COLOR)
@@ -0,0 +1,22 @@
+"""Fixtures shared by the end-to-end tests."""
+
+import tkinter as tk
+from collections.abc import Iterator
+
+import pytest
+
+
+@pytest.fixture(scope="session")
+def tk_root() -> Iterator[tk.Tk]:
+    """The one Tk root of the process, shared by every e2e module.
+
+    A second interpreter right after destroying the first fails to find
+    init.tcl on some Windows Python installs, so there is only one. Each test
+    builds its views on it and destroys their widgets afterwards.
+    """
+    try:
+        root = tk.Tk()
+    except tk.TclError:
+        pytest.skip("no display available for Tk")
+    yield root
+    root.destroy()
@@ -29,6 +29,7 @@ from flash_card.constants import (
 from flash_card.domain.entities.card import Card
 from flash_card.domain.entities.deck import Deck
 from flash_card.domain.value_objects.language_pair import LanguagePair
+from flash_card.domain.value_objects.level import Level
 from flash_card.infrastructure.persistence.csv_progress_repository import (
    CsvProgressRepository,
 )
@@ -85,18 +86,20 @@ def _shown(view: FlashCardView) -> tuple[str, str]:
    )


+def _destroy_view(root: tk.Tk, view: FlashCardView) -> None:
+    if view.flip_timer is not None:
+        root.after_cancel(view.flip_timer)
+        view.flip_timer = None
+    for widget in (view.canvas, view.wrong_button, view.right_button):
+        widget.destroy()
+
+
@pytest.fixture(scope="module")
-def window() -> Iterator[tuple[tk.Tk, FlashCardView]]:
-    # One Tk per module: a second interpreter right after destroying the
-    # first fails to find init.tcl on some Windows Python installs.
-    try:
-        root = tk.Tk()
-    except tk.TclError:
-        pytest.skip("no display available for Tk")
-    view = _build_view(root)
-    root.update_idletasks()
-    yield root, view
-    root.destroy()
+def window(tk_root: tk.Tk) -> Iterator[tuple[tk.Tk, FlashCardView]]:
+    view = _build_view(tk_root)
+    tk_root.update_idletasks()
+    yield tk_root, view
+    _destroy_view(tk_root, view)


 def test_window_has_title_background_and_padding(
@@ -306,6 +309,7 @@ def files_view(
        CsvProgressRepository(progress),
        NullSpeech(),
        _first,
+        Level(None),
    )
    view = FlashCardView(root, FlashCardViewModel(session))
    yield view, progress
@@ -0,0 +1,209 @@
+"""End-to-end tests of the level screen and the flow to the first card.
+
+@file test_level_window.py
+@brief Builds the real level screen with Tk and checks it against the wireframe.
+"""
+
+import shutil
+import tkinter as tk
+from collections.abc import Iterator, Sequence
+from pathlib import Path
+
+import pytest
+
+from flash_card.constants import (
+    BACKGROUND_COLOR,
+    CARD_WIDTH,
+    DEFAULT_DATA_DIR,
+    FINISHED_MESSAGE,
+    LEVEL_BUTTON_COLOR,
+    LEVEL_BUTTON_FONT,
+    LEVEL_SCREEN_HEIGHT,
+    LEVEL_TITLE,
+    LEVEL_TITLE_FONT,
+    PROGRESS_FILE_NAME,
+    VOCABULARY_FILE_NAME,
+    WINDOW_PADDING,
+    WINDOW_TITLE,
+)
+from flash_card.domain.entities.card import Card
+from flash_card.domain.entities.deck import Deck
+from flash_card.domain.value_objects.level import Level
+from flash_card.infrastructure.persistence.csv_progress_repository import (
+    CsvProgressRepository,
+)
+from flash_card.infrastructure.persistence.csv_vocabulary_repository import (
+    CsvVocabularyRepository,
+)
+from flash_card.infrastructure.speech.null_speech import NullSpeech
+from flash_card.presentation import app
+from flash_card.presentation.features.flash_card.flash_card_view import FlashCardView
+from flash_card.presentation.features.level.level_view import LevelView
+from flash_card.presentation.features.level.level_view_model import LevelViewModel
+
+
+def _last(cards: Sequence[Card]) -> Card:
+    return cards[-1]
+
+
+def _label(button: tk.Button) -> str:
+    return str(button.cget("text"))
+
+
+def _word(view: FlashCardView) -> str:
+    return str(view.canvas.itemcget(view.word_text_id, "text"))
+
+
+def _clear(root: tk.Tk, view: object) -> None:
+    """Destroy the widgets of whichever screen is shown, and pending timers."""
+    if isinstance(view, FlashCardView):
+        if view.flip_timer is not None:
+            root.after_cancel(view.flip_timer)
+            view.flip_timer = None
+        for widget in (view.canvas, view.wrong_button, view.right_button):
+            widget.destroy()
+    elif isinstance(view, LevelView):
+        view.destroy()
+
+
+@pytest.fixture
+def data_dir(tmp_path: Path) -> Path:
+    shutil.copy(DEFAULT_DATA_DIR / VOCABULARY_FILE_NAME, tmp_path)
+    return tmp_path
+
+
+@pytest.fixture
+def application(tk_root: tk.Tk, data_dir: Path) -> Iterator[app.Application]:
+    instance = app.Application(
+        tk_root,
+        CsvVocabularyRepository(data_dir / VOCABULARY_FILE_NAME),
+        CsvProgressRepository(data_dir / PROGRESS_FILE_NAME),
+        NullSpeech(),
+        _last,
+    )
+    yield instance
+    _clear(tk_root, instance.view)
+
+
+@pytest.fixture
+def level_view(tk_root: tk.Tk) -> Iterator[LevelView]:
+    chosen: list[Level] = []
+    view = LevelView(tk_root, LevelViewModel(), chosen.append)
+    tk_root.update_idletasks()
+    yield view
+    view.destroy()
+
+
+def test_level_screen_shows_title_with_font_on_window_background(
+    level_view: LevelView,
+) -> None:
+    label = level_view.title_label
+    assert label.cget("text") == LEVEL_TITLE == "Choose your level"
+    assert label.cget("bg") == BACKGROUND_COLOR
+    family, size, style = LEVEL_TITLE_FONT
+    assert str(label.cget("font")).split() == [family, str(size), style]
+    assert level_view.frame.cget("bg") == BACKGROUND_COLOR
+
+
+def test_level_screen_frame_has_the_card_window_size(level_view: LevelView) -> None:
+    assert level_view.frame.winfo_width() == CARD_WIDTH
+    assert level_view.frame.winfo_height() == LEVEL_SCREEN_HEIGHT
+
+
+def test_level_screen_shows_three_buttons_in_order(level_view: LevelView) -> None:
+    labels = [_label(button) for button in level_view.buttons]
+    assert labels == ["100 words", "1000 words", "All words"]
+    rows = [int(button.grid_info()["row"]) for button in level_view.buttons]
+    assert rows == [1, 2, 3]
+
+
+def test_level_buttons_have_equal_width_font_and_colour(
+    level_view: LevelView,
+) -> None:
+    widths = {int(str(button.cget("width"))) for button in level_view.buttons}
+    assert len(widths) == 1
+    family, size, style = LEVEL_BUTTON_FONT
+    for button in level_view.buttons:
+        assert str(button.cget("font")).split() == [family, str(size), style]
+        assert button.cget("bg") == LEVEL_BUTTON_COLOR
+    assert len({button.winfo_reqwidth() for button in level_view.buttons}) == 1
+
+
+def test_clicking_a_button_forwards_its_level(tk_root: tk.Tk) -> None:
+    chosen: list[Level] = []
+    view = LevelView(tk_root, LevelViewModel(), chosen.append)
+    for button in view.buttons:
+        button.invoke()
+    view.destroy()
+    assert chosen == [Level(100), Level(1000), Level(None)]
+
+
+def test_destroy_removes_the_level_frame(tk_root: tk.Tk) -> None:
+    view = LevelView(tk_root, LevelViewModel(), lambda level: None)
+    view.destroy()
+    assert not view.frame.winfo_exists()
+
+
+def test_clicking_100_words_replaces_screen_with_first_card_of_first_hundred(
+    tk_root: tk.Tk, application: app.Application, data_dir: Path
+) -> None:
+    first_hundred = (
+        CsvVocabularyRepository(data_dir / VOCABULARY_FILE_NAME).load().cards[:100]
+    )
+    application.show_level_screen()
+    level_view = application.view
+    assert isinstance(level_view, LevelView)
+
+    level_view.buttons[0].invoke()
+
+    card_view = application.view
+    assert isinstance(card_view, FlashCardView)
+    assert not level_view.frame.winfo_exists()
+    assert _word(card_view) == first_hundred[-1].front
+    assert tk_root.title() == WINDOW_TITLE
+    assert int(str(tk_root.cget("padx"))) == WINDOW_PADDING
+
+
+def test_level_whose_words_are_all_known_shows_the_finished_message(
+    application: app.Application, data_dir: Path
+) -> None:
+    cards = CsvVocabularyRepository(data_dir / VOCABULARY_FILE_NAME).load().cards
+    pair = CsvVocabularyRepository(data_dir / VOCABULARY_FILE_NAME).load().language_pair
+    CsvProgressRepository(data_dir / PROGRESS_FILE_NAME).save(Deck(pair, cards[100:]))
+    application.show_level_screen()
+    level_view = application.view
+    assert isinstance(level_view, LevelView)
+
+    level_view.buttons[0].invoke()
+
+    card_view = application.view
+    assert isinstance(card_view, FlashCardView)
+    assert _word(card_view) == FINISHED_MESSAGE
+
+
+def test_create_application_with_level_skips_the_level_screen(
+    tk_root: tk.Tk, data_dir: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    monkeypatch.setattr(tk, "Tk", lambda: tk_root)
+
+    application = app.create_application(data_dir, False, Level(100))
+
+    try:
+        assert isinstance(application.view, FlashCardView)
+        assert not [c for c in tk_root.winfo_children() if isinstance(c, tk.Frame)]
+    finally:
+        _clear(tk_root, application.view)
+
+
+def test_create_application_without_level_shows_the_level_screen(
+    tk_root: tk.Tk, data_dir: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    monkeypatch.setattr(tk, "Tk", lambda: tk_root)
+
+    application = app.create_application(data_dir, False)
+
+    try:
+        assert isinstance(application.view, LevelView)
+        assert application.root is tk_root
+    finally:
+        _clear(tk_root, application.view)
@@ -0,0 +1,124 @@
+"""Integration tests of the levels over the real word list and a progress file."""
+
+from collections.abc import Callable, Sequence
+from pathlib import Path
+
+from flash_card.application.use_cases.open_study_session import open_study_session
+from flash_card.application.use_cases.study_session import StudySession
+from flash_card.constants import (
+    DEFAULT_DATA_DIR,
+    PROGRESS_FILE_NAME,
+    VOCABULARY_FILE_NAME,
+)
+from flash_card.domain.entities.card import Card
+from flash_card.domain.value_objects.level import Level
+from flash_card.infrastructure.persistence.csv_progress_repository import (
+    CsvProgressRepository,
+)
+from flash_card.infrastructure.persistence.csv_vocabulary_repository import (
+    CsvVocabularyRepository,
+)
+from flash_card.infrastructure.speech.null_speech import NullSpeech
+
+VOCABULARY = CsvVocabularyRepository(DEFAULT_DATA_DIR / VOCABULARY_FILE_NAME)
+ALL_CARDS = VOCABULARY.load().cards
+FIRST_HUNDRED = ALL_CARDS[:100]
+
+Choose = Callable[[Sequence[Card]], Card]
+
+
+def _first(cards: Sequence[Card]) -> Card:
+    return cards[0]
+
+
+def _open(folder: Path, level: Level, choose: Choose = _first) -> StudySession:
+    return open_study_session(
+        VOCABULARY,
+        CsvProgressRepository(folder / PROGRESS_FILE_NAME),
+        NullSpeech(),
+        choose,
+        level,
+    )
+
+
+def _learn_to_the_end(session: StudySession) -> list[Card]:
+    """Mark every drawn card known until the session is finished."""
+    drawn: list[Card] = []
+    while (card := session.next_card()) is not None:
+        drawn.append(Card(card.front_text, card.back_text))
+        session.mark_known()
+    return drawn
+
+
+def _saved(folder: Path) -> tuple[Card, ...]:
+    loaded = CsvProgressRepository(folder / PROGRESS_FILE_NAME).load()
+    assert loaded is not None
+    return loaded.cards
+
+
+def test_word_list_has_one_thousand_distinct_words() -> None:
+    assert len(ALL_CARDS) == 1000
+    assert len(set(ALL_CARDS)) == 1000
+
+
+def test_level_100_draws_exactly_the_first_hundred_words(tmp_path: Path) -> None:
+    drawn = _learn_to_the_end(_open(tmp_path, Level(100)))
+    assert len(drawn) == 100
+    assert len(set(drawn)) == 100
+    assert set(drawn) == set(FIRST_HUNDRED)
+
+
+def test_level_100_keeps_the_other_words_in_order_in_the_file(
+    tmp_path: Path,
+) -> None:
+    _learn_to_the_end(_open(tmp_path, Level(100)))
+    assert _saved(tmp_path) == ALL_CARDS[100:]
+    assert len(_saved(tmp_path)) == 900
+
+
+def test_level_all_after_level_100_draws_only_the_other_words(
+    tmp_path: Path,
+) -> None:
+    _learn_to_the_end(_open(tmp_path, Level(100)))
+    drawn = _learn_to_the_end(_open(tmp_path, Level(None)))
+    assert len(drawn) == 900
+    assert set(drawn) == set(ALL_CARDS[100:])
+    assert not set(drawn) & set(FIRST_HUNDRED)
+
+
+def test_level_1000_equals_all_words_for_this_list(tmp_path: Path) -> None:
+    session_1000 = _open(tmp_path, Level(1000))
+    session_all = _open(tmp_path, Level(None))
+    assert session_1000.remaining_in_level == 1000
+    assert session_all.remaining_in_level == 1000
+
+
+def test_words_known_at_level_100_stay_known_at_level_1000(tmp_path: Path) -> None:
+    first = _open(tmp_path, Level(100))
+    shown = first.next_card()
+    assert shown is not None
+    first.mark_known()
+    session = _open(tmp_path, Level(1000))
+    assert session.remaining == 999
+    assert session.remaining_in_level == 999
+
+
+def test_marking_known_at_level_100_keeps_words_outside_the_level(
+    tmp_path: Path,
+) -> None:
+    session = _open(tmp_path, Level(100))
+    session.next_card()
+    session.mark_known()
+    saved = _saved(tmp_path)
+    assert len(saved) == 999
+    assert saved[-900:] == ALL_CARDS[100:]
+
+
+def test_level_whose_words_are_all_known_is_finished_at_once(
+    tmp_path: Path,
+) -> None:
+    _learn_to_the_end(_open(tmp_path, Level(100)))
+    session = _open(tmp_path, Level(100))
+    assert session.remaining_in_level == 0
+    assert session.remaining == 900
+    assert session.next_card() is None
@@ -6,6 +6,7 @@ from pathlib import Path
 from flash_card.application.use_cases.open_study_session import open_study_session
 from flash_card.application.use_cases.study_session import StudySession
 from flash_card.domain.entities.card import Card
+from flash_card.domain.value_objects.level import Level
 from flash_card.infrastructure.persistence.csv_progress_repository import (
    CsvProgressRepository,
 )
@@ -36,6 +37,7 @@ def _open(folder: Path, choose: Choose = _first) -> StudySession:
        CsvProgressRepository(folder / "words_to_learn.csv"),
        NullSpeech(),
        choose,
+        Level(None),
    )


@@ -10,6 +10,7 @@ from flash_card.constants import (
 )
 from flash_card.domain.entities.card import Card
 from flash_card.domain.value_objects.language_pair import LanguagePair
+from flash_card.domain.value_objects.level import Level
 from flash_card.infrastructure.persistence.csv_progress_repository import (
    CsvProgressRepository,
 )
@@ -34,7 +35,11 @@ def test_open_study_session_yields_card_from_real_file(tmp_path: Path) -> None:
    cards = repository.load().cards
    progress = CsvProgressRepository(tmp_path / PROGRESS_FILE_NAME)
    session = open_study_session(
-        repository, progress, NullSpeech(), lambda candidates: candidates[0]
+        repository,
+        progress,
+        NullSpeech(),
+        lambda candidates: candidates[0],
+        Level(None),
    )

    card = session.next_card()
@@ -11,6 +11,7 @@ from flash_card.domain.entities.deck import Deck
 from flash_card.domain.entities.vocabulary import Vocabulary
 from flash_card.domain.errors.vocabulary_errors import VocabularyNotFoundError
 from flash_card.domain.value_objects.language_pair import LanguagePair
+from flash_card.domain.value_objects.level import Level


 class _FakeRepository:
@@ -52,7 +53,7 @@ def test_open_study_session_builds_session_over_all_cards() -> None:
    repository = _FakeRepository(Vocabulary(LanguagePair("ms", "da"), cards))

    session = open_study_session(
-        repository, _RecordingProgress(), _SilentSpeech(), _first
+        repository, _RecordingProgress(), _SilentSpeech(), _first, Level(None)
    )

    assert session.remaining == 2
@@ -62,7 +63,11 @@ def test_open_study_session_builds_session_over_all_cards() -> None:
 def test_open_study_session_propagates_repository_error() -> None:
    with pytest.raises(VocabularyNotFoundError):
        open_study_session(
-            _MissingRepository(), _RecordingProgress(), _SilentSpeech(), _first
+            _MissingRepository(),
+            _RecordingProgress(),
+            _SilentSpeech(),
+            _first,
+            Level(None),
        )


@@ -72,21 +77,83 @@ def test_open_study_session_prefers_saved_progress() -> None:
    progress = _RecordingProgress(Deck(pair, [Card("kau", "du")]))

    session = open_study_session(
-        _FakeRepository(vocabulary), progress, _SilentSpeech(), _first
+        _FakeRepository(vocabulary), progress, _SilentSpeech(), _first, Level(None)
    )

    assert session.remaining == 1
    assert session.next_card() == CardDto("kau", "du", "ms", "da")


-def test_open_study_session_skips_vocabulary_when_progress_exists() -> None:
-    progress = _RecordingProgress(Deck(LanguagePair("ms", "da"), []))
+PAIR = LanguagePair("ms", "da")
+THREE = (Card("aku", "mig"), Card("kau", "du"), Card("ini", "denne"))

-    session = open_study_session(
-        _MissingRepository(), progress, _SilentSpeech(), _first
+
+def _last(cards: Sequence[Card]) -> Card:
+    return cards[-1]
+
+
+def test_open_study_session_propagates_error_when_list_is_missing() -> None:
+    progress = _RecordingProgress(Deck(PAIR, []))
+
+    with pytest.raises(VocabularyNotFoundError):
+        open_study_session(
+            _MissingRepository(), progress, _SilentSpeech(), _first, Level(None)
        )

-    assert session.remaining == 0
+
+def test_open_study_session_limits_draws_to_first_cards_of_level() -> None:
+    session = open_study_session(
+        _FakeRepository(Vocabulary(PAIR, THREE)),
+        _RecordingProgress(),
+        _SilentSpeech(),
+        _last,
+        Level(2),
+    )
+
+    assert session.remaining == 3
+    assert session.remaining_in_level == 2
+    assert session.next_card() == CardDto("kau", "du", "ms", "da")
+
+
+def test_open_study_session_uses_vocabulary_order_with_saved_progress() -> None:
+    progress = _RecordingProgress(Deck(PAIR, [THREE[2], THREE[1]]))
+    session = open_study_session(
+        _FakeRepository(Vocabulary(PAIR, THREE)),
+        progress,
+        _SilentSpeech(),
+        _first,
+        Level(2),
+    )
+
+    assert session.remaining == 2
+    assert session.remaining_in_level == 1
+    assert session.next_card() == CardDto("kau", "du", "ms", "da")
+
+
+def test_open_study_session_is_finished_when_level_has_nothing_left() -> None:
+    progress = _RecordingProgress(Deck(PAIR, [THREE[1]]))
+    session = open_study_session(
+        _FakeRepository(Vocabulary(PAIR, THREE)),
+        progress,
+        _SilentSpeech(),
+        _first,
+        Level(1),
+    )
+
+    assert session.next_card() is None
+    assert session.remaining == 1
+
+
+def test_open_study_session_uses_all_cards_when_level_exceeds_list() -> None:
+    session = open_study_session(
+        _FakeRepository(Vocabulary(PAIR, THREE)),
+        _RecordingProgress(),
+        _SilentSpeech(),
+        _first,
+        Level(1000),
+    )
+
+    assert session.remaining_in_level == 3


 def test_open_study_session_never_saves_on_start() -> None:
@@ -94,7 +161,7 @@ def test_open_study_session_never_saves_on_start() -> None:
    progress = _RecordingProgress()

    session = open_study_session(
-        _FakeRepository(vocabulary), progress, _SilentSpeech(), _first
+        _FakeRepository(vocabulary), progress, _SilentSpeech(), _first, Level(None)
    )
    session.next_card()

@@ -1,6 +1,7 @@
 """Tests of the StudySession use case."""

-from collections.abc import Sequence
+import itertools
+from collections.abc import Callable, Sequence

 from flash_card.application.dto.card_dto import CardDto
 from flash_card.application.use_cases.study_session import StudySession
@@ -150,3 +151,94 @@ def test_mark_known_does_not_speak_but_following_draw_does() -> None:
    assert speech.spoken == [("aku", "ms")]
    session.next_card()
    assert speech.spoken == [("aku", "ms"), ("kau", "ms")]
+
+
+C = Card("ini", "denne")
+D = Card("itu", "den")
+
+
+def _level_session(
+    deck_cards: Sequence[Card],
+    level_cards: Sequence[Card] | None,
+    choose: Callable[[Sequence[Card]], Card] = _first,
+    progress: _RecordingProgress | None = None,
+    speech: _RecordingSpeech | None = None,
+) -> StudySession:
+    store = _RecordingProgress() if progress is None else progress
+    voice = _RecordingSpeech() if speech is None else speech
+    return StudySession(Deck(PAIR, deck_cards), choose, store, voice, level_cards)
+
+
+def _last(cards: Sequence[Card]) -> Card:
+    return cards[-1]
+
+
+def test_next_card_draws_only_cards_of_the_level_over_many_draws() -> None:
+    chooser = itertools.cycle([_first, _last])
+    session = _level_session(
+        [A, B, C, D], [A, B], choose=lambda cards: next(chooser)(cards)
+    )
+    drawn = {session.next_card() for _ in range(40)}
+    assert drawn == {
+        CardDto("aku", "mig", "ms", "da"),
+        CardDto("kau", "du", "ms", "da"),
+    }
+
+
+def test_next_card_never_draws_a_card_outside_the_level() -> None:
+    session = _level_session([A, B, C], [B], choose=_last)
+    for _ in range(5):
+        assert session.next_card() == CardDto("kau", "du", "ms", "da")
+
+
+def test_remaining_counts_whole_deck_and_remaining_in_level_counts_level() -> None:
+    session = _level_session([A, B, C], [A, B])
+    assert session.remaining == 3
+    assert session.remaining_in_level == 2
+
+
+def test_remaining_in_level_counts_whole_deck_without_level() -> None:
+    assert _level_session([A, B, C], None).remaining_in_level == 3
+
+
+def test_remaining_in_level_ignores_level_cards_missing_from_deck() -> None:
+    assert _level_session([B], [A, B]).remaining_in_level == 1
+
+
+def test_mark_known_keeps_cards_outside_the_level_and_saves_whole_deck() -> None:
+    progress = _RecordingProgress()
+    session = _level_session([A, B, C], [A], progress=progress)
+    session.next_card()
+    session.mark_known()
+    assert progress.saved == [(B, C)]
+    assert session.remaining == 2
+    assert session.remaining_in_level == 0
+
+
+def test_next_card_is_none_when_nothing_of_the_level_is_left() -> None:
+    session = _level_session([B, C], [A])
+    assert session.next_card() is None
+    assert session.current_card is None
+
+
+def test_session_finishes_after_last_card_of_level_is_known() -> None:
+    session = _level_session([A, B], [A])
+    session.next_card()
+    session.mark_known()
+    assert session.next_card() is None
+    assert session.remaining == 1
+
+
+def test_next_card_speaks_only_cards_of_the_level() -> None:
+    speech = _RecordingSpeech()
+    session = _level_session([A, B, C], [B], choose=_last, speech=speech)
+    session.next_card()
+    session.next_card()
+    assert speech.spoken == [("kau", "ms"), ("kau", "ms")]
+
+
+def test_next_card_speaks_nothing_when_level_is_finished() -> None:
+    speech = _RecordingSpeech()
+    session = _level_session([B], [A], speech=speech)
+    session.next_card()
+    assert speech.spoken == []
@@ -71,3 +71,47 @@ def test_remove_raises_when_card_is_absent() -> None:
 def test_is_empty_reflects_cards() -> None:
    assert Deck(PAIR, []).is_empty
    assert not Deck(PAIR, [A]).is_empty
+
+
+def test_within_returns_deck_cards_found_in_given_cards_in_deck_order() -> None:
+    c = Card("ini", "denne")
+    deck = Deck(PAIR, [A, B, c])
+    assert deck.within([c, A]) == (A, c)
+
+
+def test_within_respects_multiplicity() -> None:
+    deck = Deck(PAIR, [A, A, B])
+    assert deck.within([A, B]) == (A, B)
+    assert deck.within([A, A, A]) == (A, A)
+
+
+def test_within_returns_empty_when_no_card_matches() -> None:
+    assert Deck(PAIR, [A]).within([B]) == ()
+
+
+def test_within_does_not_change_the_deck() -> None:
+    deck = Deck(PAIR, [A, B])
+    deck.within([A])
+    assert deck.cards == (A, B)
+
+
+def test_draw_chooses_among_given_candidates_only() -> None:
+    deck = Deck(PAIR, [A, B])
+    assert deck.draw(_first, among=[B]) == B
+
+
+def test_draw_among_avoids_excluded_when_another_candidate_exists() -> None:
+    c = Card("ini", "denne")
+    deck = Deck(PAIR, [A, B, c])
+    assert deck.draw(_first, exclude=B, among=[B, c]) == c
+
+
+def test_draw_among_returns_excluded_when_it_is_the_only_candidate() -> None:
+    deck = Deck(PAIR, [A, B])
+    assert deck.draw(_first, exclude=B, among=[B]) == B
+
+
+def test_draw_among_empty_raises_empty_deck_error() -> None:
+    deck = Deck(PAIR, [A, B])
+    with pytest.raises(EmptyDeckError):
+        deck.draw(_first, among=[])
@@ -0,0 +1,42 @@
+"""Tests of the Level value object."""
+
+import dataclasses
+
+import pytest
+
+from flash_card.domain.entities.card import Card
+from flash_card.domain.value_objects.level import Level
+
+CARDS = tuple(Card(f"w{i}", f"v{i}") for i in range(1000))
+
+
+@pytest.mark.parametrize("limit", [0, -1, -100])
+def test_level_refuses_limit_not_above_zero(limit: int) -> None:
+    with pytest.raises(ValueError, match="above zero"):
+        Level(limit)
+
+
+def test_level_accepts_no_limit() -> None:
+    assert Level(None).limit is None
+
+
+def test_level_is_frozen() -> None:
+    level = Level(100)
+    with pytest.raises(dataclasses.FrozenInstanceError):
+        level.limit = 5  # type: ignore[misc]
+
+
+def test_first_returns_first_hundred_cards_when_limit_is_100() -> None:
+    assert Level(100).first(CARDS) == CARDS[:100]
+
+
+def test_first_returns_all_cards_when_there_is_no_limit() -> None:
+    assert Level(None).first(CARDS) == CARDS
+
+
+def test_first_returns_all_cards_when_limit_exceeds_the_list() -> None:
+    assert Level(5000).first(CARDS) == CARDS
+
+
+def test_first_returns_empty_when_list_is_empty() -> None:
+    assert Level(100).first(()) == ()
@@ -1,45 +1,243 @@
 """Tests of the entry point.

@file test_app.py
-@brief Checks main of flash_card.presentation.app without opening a window.
+@brief Checks main and Application of flash_card.presentation.app without Tk.
 """

+import tkinter as tk
+from collections.abc import Callable, Sequence
 from pathlib import Path
+from typing import cast

 import pytest

+from flash_card.domain.entities.card import Card
+from flash_card.domain.entities.deck import Deck
+from flash_card.domain.entities.vocabulary import Vocabulary
+from flash_card.domain.errors.vocabulary_errors import VocabularyNotFoundError
+from flash_card.domain.value_objects.language_pair import LanguagePair
+from flash_card.domain.value_objects.level import Level
 from flash_card.presentation import app
+from flash_card.presentation.features.flash_card.flash_card_view_model import (
+    FlashCardViewModel,
+)
+from flash_card.presentation.features.level.level_view_model import LevelViewModel
+
+PAIR = LanguagePair("ms", "da")
+CARDS = (Card("aku", "mig"), Card("kau", "du"), Card("ini", "denne"))


 class _FakeRoot:
    def __init__(self) -> None:
        self.mainloop_calls = 0
+        self.destroyed = False

    def mainloop(self) -> None:
        self.mainloop_calls += 1

+    def destroy(self) -> None:
+        self.destroyed = True
+
+
+class _FakeApplication:
+    def __init__(self, root: _FakeRoot, exit_code: int = 0) -> None:
+        self.root = root
+        self.exit_code = exit_code
+
+
+class _FakeLevelView:
+    def __init__(
+        self,
+        root: object,
+        view_model: LevelViewModel,
+        on_choose: Callable[[Level], None],
+    ) -> None:
+        self.view_model = view_model
+        self.on_choose = on_choose
+        self.destroyed = False
+
+    def destroy(self) -> None:
+        self.destroyed = True
+
+
+class _FakeFlashCardView:
+    def __init__(self, root: object, view_model: FlashCardViewModel) -> None:
+        self.view_model = view_model
+
+
+class _Vocabulary:
+    def __init__(self, error: bool = False) -> None:
+        self._error = error
+
+    def load(self) -> Vocabulary:
+        if self._error:
+            raise VocabularyNotFoundError("word list not found")
+        return Vocabulary(PAIR, CARDS)
+
+
+class _Progress:
+    def load(self) -> Deck | None:
+        return None
+
+    def save(self, deck: Deck) -> None:
+        return None
+
+
+class _Speech:
+    def speak(self, text: str, language_code: str) -> None:
+        return None
+
+
+def _first(cards: Sequence[Card]) -> Card:
+    return cards[0]
+
+
+@pytest.fixture
+def fake_views(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(app, "LevelView", _FakeLevelView)
+    monkeypatch.setattr(app, "FlashCardView", _FakeFlashCardView)
+
+
+def _application(root: _FakeRoot, error: bool = False) -> app.Application:
+    return app.Application(
+        cast(tk.Tk, root), _Vocabulary(error), _Progress(), _Speech(), _first
+    )
+
+
+def _recording_create(
+    seen: list[Level | None], exit_code: int = 0
+) -> Callable[..., _FakeApplication]:
+    def fake_create(
+        data_dir: Path | None = None,
+        speech_enabled: bool = True,
+        level: Level | None = None,
+    ) -> _FakeApplication:
+        seen.append(level)
+        return _FakeApplication(_FakeRoot(), exit_code)
+
+    return fake_create
+

 def test_main_runs_main_loop_and_returns_zero(
    monkeypatch: pytest.MonkeyPatch, tmp_path: Path
 ) -> None:
    root = _FakeRoot()
-    seen: list[Path | None] = []
+    seen: list[tuple[Path | None, Level | None]] = []

    def fake_create(
-        data_dir: Path | None = None, speech_enabled: bool = True
-    ) -> tuple[_FakeRoot, object]:
-        seen.append(data_dir)
-        return root, object()
+        data_dir: Path | None = None,
+        speech_enabled: bool = True,
+        level: Level | None = None,
+    ) -> _FakeApplication:
+        seen.append((data_dir, level))
+        return _FakeApplication(root)

    monkeypatch.setattr(app, "create_application", fake_create)

    assert app.main(["--data-dir", str(tmp_path)]) == 0
    assert root.mainloop_calls == 1
-    assert seen == [tmp_path]
+    assert seen == [(tmp_path, None)]


-def test_main_returns_one_when_word_list_is_missing(
+def test_main_returns_one_when_word_list_is_missing_with_level(
    tmp_path: Path, caplog: pytest.LogCaptureFixture
 ) -> None:
-    assert app.main(["--data-dir", str(tmp_path / "nowhere")]) == 1
+    code = app.main(["--data-dir", str(tmp_path / "nowhere"), "--level", "all"])
+    assert code == 1
    assert "word list not found" in caplog.text
+
+
+@pytest.mark.parametrize(
+    ("value", "limit"), [("100", 100), ("1000", 1000), ("all", None)]
+)
+def test_main_passes_level_of_level_option(
+    monkeypatch: pytest.MonkeyPatch, tmp_path: Path, value: str, limit: int | None
+) -> None:
+    seen: list[Level | None] = []
+    monkeypatch.setattr(app, "create_application", _recording_create(seen))
+
+    assert app.main(["--data-dir", str(tmp_path), "--level", value]) == 0
+    assert seen == [Level(limit)]
+
+
+def test_main_refuses_unknown_level(capsys: pytest.CaptureFixture[str]) -> None:
+    with pytest.raises(SystemExit) as raised:
+        app.main(["--level", "50"])
+    assert raised.value.code == 2
+    assert "invalid level" in capsys.readouterr().err
+
+
+def test_main_returns_exit_code_of_application(
+    monkeypatch: pytest.MonkeyPatch, tmp_path: Path
+) -> None:
+    seen: list[Level | None] = []
+    monkeypatch.setattr(app, "create_application", _recording_create(seen, 1))
+
+    assert app.main(["--data-dir", str(tmp_path)]) == 1
+
+
+def test_application_shows_no_view_before_a_screen_is_requested(
+    fake_views: None,
+) -> None:
+    assert _application(_FakeRoot()).view is None
+
+
+def test_show_level_screen_offers_the_three_levels(fake_views: None) -> None:
+    application = _application(_FakeRoot())
+    application.show_level_screen()
+    view = application.view
+    assert isinstance(view, _FakeLevelView)
+    assert [c.level.limit for c in view.view_model.choices] == [100, 1000, None]
+
+
+def test_choosing_a_level_replaces_level_screen_by_card_screen(
+    fake_views: None,
+) -> None:
+    application = _application(_FakeRoot())
+    application.show_level_screen()
+    level_view = application.view
+    assert isinstance(level_view, _FakeLevelView)
+
+    level_view.on_choose(Level(100))
+
+    assert level_view.destroyed
+    assert isinstance(application.view, _FakeFlashCardView)
+    assert application.exit_code == 0
+
+
+def test_start_draws_the_first_card_of_the_level(fake_views: None) -> None:
+    application = _application(_FakeRoot())
+    application.start(Level(2))
+    view = application.view
+    assert isinstance(view, _FakeFlashCardView)
+    assert view.view_model.text == "aku"
+
+
+def test_start_without_level_screen_shows_card_screen_only(
+    fake_views: None,
+) -> None:
+    application = _application(_FakeRoot())
+    application.start(Level(None))
+    assert isinstance(application.view, _FakeFlashCardView)
+
+
+def test_choosing_a_level_closes_window_with_code_one_on_vocabulary_error(
+    fake_views: None, caplog: pytest.LogCaptureFixture
+) -> None:
+    root = _FakeRoot()
+    application = _application(root, error=True)
+    application.show_level_screen()
+    level_view = application.view
+    assert isinstance(level_view, _FakeLevelView)
+
+    level_view.on_choose(Level(100))
+
+    assert root.destroyed
+    assert application.exit_code == 1
+    assert "word list not found" in caplog.text
+
+
+def test_start_raises_vocabulary_error_for_the_caller(fake_views: None) -> None:
+    application = _application(_FakeRoot(), error=True)
+    with pytest.raises(VocabularyNotFoundError):
+        application.start(Level(100))
@@ -6,6 +6,7 @@ from pathlib import Path

 import pytest

+from flash_card.domain.value_objects.level import Level
 from flash_card.infrastructure.speech.gtts_speech import GttsSpeech
 from flash_card.infrastructure.speech.null_speech import NullSpeech
 from flash_card.presentation import app
@@ -16,6 +17,11 @@ class _FakeRoot:
        return None


+class _FakeApplication:
+    root = _FakeRoot()
+    exit_code = 0
+
+
 def _found(name: str) -> ModuleSpec:
    return ModuleSpec(name, None)

@@ -58,10 +64,12 @@ def test_main_passes_speech_disabled_with_no_speech_option(
    seen: list[bool] = []

    def fake_create(
-        data_dir: Path | None = None, speech_enabled: bool = True
-    ) -> tuple[_FakeRoot, object]:
+        data_dir: Path | None = None,
+        speech_enabled: bool = True,
+        level: Level | None = None,
+    ) -> _FakeApplication:
        seen.append(speech_enabled)
-        return _FakeRoot(), object()
+        return _FakeApplication()

    monkeypatch.setattr(app, "create_application", fake_create)

@@ -0,0 +1,46 @@
+"""Tests of the level view model."""
+
+import dataclasses
+
+import pytest
+
+from flash_card.constants import LEVEL_TITLE
+from flash_card.domain.value_objects.level import Level
+from flash_card.presentation.features.level import level_view_model
+from flash_card.presentation.features.level.level_view_model import (
+    LevelChoice,
+    LevelViewModel,
+)
+
+
+def test_title_is_choose_your_level() -> None:
+    assert LevelViewModel().title == LEVEL_TITLE == "Choose your level"
+
+
+def test_choices_are_three_levels_in_order() -> None:
+    choices = LevelViewModel().choices
+    assert [choice.label for choice in choices] == [
+        "100 words",
+        "1000 words",
+        "All words",
+    ]
+    assert [choice.level for choice in choices] == [
+        Level(100),
+        Level(1000),
+        Level(None),
+    ]
+
+
+def test_choices_come_from_given_choices() -> None:
+    model = LevelViewModel([("Ten", 10)])
+    assert model.choices == (LevelChoice("Ten", Level(10)),)
+
+
+def test_choice_is_frozen() -> None:
+    choice = LevelChoice("Ten", Level(10))
+    with pytest.raises(dataclasses.FrozenInstanceError):
+        choice.label = "x"  # type: ignore[misc]
+
+
+def test_view_model_module_does_not_import_tkinter() -> None:
+    assert "tkinter" not in vars(level_view_model)
@@ -1,8 +1,8 @@
 """Capture screenshots of the running flash card window.

@file capture_screenshots.py
-@brief Starts the program, grabs the front of a card and, three seconds later,
-the back, and saves both as PNG files.
+@brief Starts the program, grabs the level screen, the front of a card and,
+three seconds later, the back, and saves them as PNG files.

 The tool needs Pillow (`python -m pip install -e ".[tools]"`), which is a tool
 dependency only and not a dependency of the program. It grabs the screen area
@@ -31,7 +31,11 @@ from flash_card.constants import (  # noqa: E402
    PROJECT_ROOT,
    VOCABULARY_FILE_NAME,
 )
+from flash_card.domain.value_objects.level import Level  # noqa: E402
 from flash_card.presentation.app import create_application  # noqa: E402
+from flash_card.presentation.features.flash_card.flash_card_view import (  # noqa: E402
+    FlashCardView,
+)

 ## @brief Default folder of the screenshots.
 DEFAULT_OUTPUT_DIR = PROJECT_ROOT / "docs" / "images"
@@ -60,21 +64,37 @@ def _pump(root: object, seconds: float) -> None:


 def capture(output_dir: Path) -> list[Path]:
-    """Capture the front and the back of one card.
+    """Capture the level screen, then the front and the back of one card.

-    @param output_dir Folder to write `screenshot-front.png` and
-        `screenshot-back.png` into.
-    @return The two files written, front first.
+    @param output_dir Folder to write `screenshot-level.png`,
+        `screenshot-front.png` and `screenshot-back.png` into.
+    @return The three files written: level, front, back.
    """
    _make_dpi_aware()
    output_dir.mkdir(parents=True, exist_ok=True)
    random.seed(SCREENSHOT_SEED)
+    level = output_dir / "screenshot-level.png"
+    front = output_dir / "screenshot-front.png"
+    back = output_dir / "screenshot-back.png"
    with tempfile.TemporaryDirectory() as data_dir:
        shutil.copy(DEFAULT_DATA_DIR / VOCABULARY_FILE_NAME, data_dir)
-        root, view = create_application(Path(data_dir), speech_enabled=False)
+        application = create_application(Path(data_dir), speech_enabled=False)
+        root = application.root
        root.lift()
        root.attributes("-topmost", True)
        _pump(root, SETTLE_SECONDS)
+        window_box = (
+            root.winfo_rootx(),
+            root.winfo_rooty(),
+            root.winfo_rootx() + root.winfo_width(),
+            root.winfo_rooty() + root.winfo_height(),
+        )
+        ImageGrab.grab(bbox=window_box).save(level)
+        application.start(Level(None))
+        view = application.view
+        if not isinstance(view, FlashCardView):
+            raise RuntimeError("the card window did not open")
+        _pump(root, SETTLE_SECONDS)
        box = (
            view.canvas.winfo_rootx() - 20,
            view.canvas.winfo_rooty() - 20,
@@ -84,13 +104,11 @@ def capture(output_dir: Path) -> list[Path]:
            + view.right_button.winfo_height()
            + 20,
        )
-        front = output_dir / "screenshot-front.png"
-        back = output_dir / "screenshot-back.png"
        ImageGrab.grab(bbox=box).save(front)
        _pump(root, FLIP_DELAY_MS / 1000 + SETTLE_SECONDS)
        ImageGrab.grab(bbox=box).save(back)
        root.destroy()
-    return [front, back]
+    return [level, front, back]


 def main(argv: Sequence[str] | None = None) -> int:
Author	SHA1	Message	Date
Tirsvad	d37f165c19	Resolve pending commit links of the MIL-009 plan	2026-10-15 00:17:31 +08:00
Tirsvad	768ca344c9	MIL-009 plan and documents: executable with PyInstaller Milestone MIL-009, story US-001.07, ADR-0004, and the amended BC-001, DCD-001 and PP-001. Review records RC-042 to RC-047, all Go (autonomous mode, not independent). Task: MIL-009#1	2026-10-15 00:17:27 +08:00
Tirsvad	d5873748cf	Merge pull request 'MIL-008: Choose level' (#54 ) from mil-008-choose-level into main TirSystem/github-action: Sync GitHub mirror metadata / sync-metadata (push) Successful in 6s Details	2026-10-14 17:58:52 +02:00
Tirsvad	6745f98c27	Resolve pending commit links of MIL-008	2026-10-14 23:58:46 +08:00
Tirsvad	f45d6036e4	MIL-008: Choose level Review record RC-041 (Go, autonomous mode, not independent). Task: MIL-008#1 Task: MIL-008#2 Task: MIL-008#3 Task: MIL-008#4 Task: MIL-008#5 Task: MIL-008#6 Task: MIL-008#7 Task: MIL-008#8	2026-10-14 23:58:31 +08:00
Tirsvad	ee9c531ea6	Merge pull request 'MIL-008 plan and design: choose level' (#45 ) from plan-mil-008-level into main TirSystem/github-action: Sync GitHub mirror metadata / sync-metadata (push) Successful in 9s Details	2026-10-14 17:45:16 +02:00
				`@@ -0,0 +1 @@`
				`"""The level feature: its view and view model."""`
				`@@ -0,0 +1 @@`
				`"""Code shared by the screens of the presentation layer."""`