Name of Thrones

Name of Thrones is an unofficial, noncommercial fan quiz. Choose one of three difficulty levels, enter names to reveal character cards, track progress, and filter the results by house or group.

Gameplay

A guess must match a full character name or a listed alias. Matching is exact after normalizing case, outside and repeated spaces, common accent marks, apostrophes, hyphens, and punctuation. Partial names and fuzzy guesses do not match. If one accepted alias belongs to more than one record, all matching records are revealed. Selecting a revealed card opens that character on A Wiki of Ice and Fire in a new tab.

The dictation helper is on by default. A close misspelling shows an underlined suggestion below the input. Selecting it uses the corrected spelling, reveals the matching card, and updates progress. The helper can be turned off, and that choice is saved in the browser.

The Options menu keeps the input area clean and stores three preferences: the dictation helper, dark mode, and automatic scrolling to a revealed card. All three options are on by default.

The three levels are:

• Newcomer: exactly 40 core television characters.
• Fan: exactly 250 named characters who appear on screen: 235 imported records plus all 15 TV-only records. It includes the Newcomer set plus recurring and credited minor characters.
• Expert: 1,015 current combined records: 1,000 imported book records plus 15 manually curated, true TV-only records.

These are curated quiz lists, not complete archives. In particular, Expert is limited to the current 1,000-record book import and 15 TV-only additions. It does not contain every named character from A Song of Ice and Fire or every character shown in Game of Thrones.

Found characters count in every level that includes them, while each level keeps its own timer and house filter in the browser's local storage. Change level keeps all saved progress and times. Reset game clears progress from every level. Progress saved by the old version 1 game is shared with every matching level when the new progress format is first saved successfully.

Release data and limits

The 1,000 imported records in data/characters.json have at least one book reference. The import excludes blank or clearly unnamed records and known TV-only, game-only, semi-canon, and The Ice Dragon markers.

The preferred importer source is A Wiki of Ice and Fire. During creation of the checked release, A Wiki of Ice and Fire returned HTTP 403, so this release uses An API of Ice and Fire instead. This fallback does not prove complete coverage of Fire & Blood. The API does not state a separate data license, so its factual data must not be treated as MIT-licensed project code.

The TV quiz data uses character names, short facts, and source links from the Wikipedia List of Game of Thrones characters, An API of Ice and Fire documentation, and the community-run Game of Thrones Wiki. It does not copy descriptive article prose or wiki images. Wikipedia text is available under Creative Commons Attribution-ShareAlike 4.0. The API does not state a separate data license and is used for TV roster mapping and reference. The Game of Thrones Wiki links to Fandom's license notice.

The level release adds these data files:

• data/show-characters.json contains the 15 manually curated TV-only records.
• data/levels.json defines the three levels, their source links, and their exact rosters or combined-record rule.
• data/character-overrides.json adds reviewed TV aliases to matching imported book records without changing the imported source data.

At runtime, the site combines data/characters.json and data/show-characters.json, applies the alias overrides, and then builds each level roster.

Run locally

The site has no build step. From the repository root, run:

python3 -m http.server

Then open http://localhost:8000/. Do not open index.html directly because the browser must fetch the JSON data over HTTP.

Checks

The checks use only Python and Node.js built-in tools:

python3 -m unittest discover -s tests -p 'test_*.py'
python3 scripts/validate_data.py
npm test

No OpenAI or Pillow package is needed for tests, data validation, or a portrait dry run.

Portrait pipeline

The checked project does not claim that generated portraits exist. House placeholders are used unless a future portrait is generated and approved.

The portrait pipeline now reads the same combined set: imported book records from data/characters.json and TV-only records from data/show-characters.json. It keeps one combined data/portrait-manifest.json. When a portrait is approved or rejected, the pipeline updates the record in the data file that owns that character. Nothing is approved automatically.

Preview a small batch without an API key, network request, OpenAI package, or Pillow package:

python3 scripts/generate_portraits.py --limit 5

Generation costs money and needs a separate local environment with the OpenAI SDK and Pillow:

python3 -m venv .venv
. .venv/bin/activate
python3 -m pip install openai Pillow
test -n "${OPENAI_API_KEY:-}"
python3 scripts/generate_portraits.py \
  --execute \
  --limit 5 \
  --confirm "GENERATE PORTRAITS"
unset OPENAI_API_KEY

Set OPENAI_API_KEY in the process environment with a trusted secret manager before running these commands. Never put the key in a command argument, repository file, screenshot, issue, chat, or commit. Never print it.

Generation never approves an image. Review each generated file for quality, safety, unwanted text, copyrighted designs, and resemblance to a real actor. Then approve or reject it by its character ID:

python3 scripts/generate_portraits.py --approve character-api-339
python3 scripts/generate_portraits.py --reject character-api-339

Approval checks that the file is a safe 512 by 512 WebP image, records the review in the manifest, and publishes its path in data/characters.json. Rejection clears the published path but keeps the image for later review.

AI-generated portraits, if added later, are artistic interpretations. They may be inaccurate or biased, are not canonical, and are not official art. They must not copy wiki images, official art, franchise designs, or actor likenesses.

Attribution and rights

See ATTRIBUTION.md, THIRD_PARTY_NOTICES.md, and LICENSE-CODE.md. The MIT license covers only original project code. It does not cover datasets, source material, franchise rights, or generated assets.

This project is not affiliated with or endorsed by George R. R. Martin, HBO, or their publishers, licensees, or rights holders. Names and story material remain the property of their respective owners. No wiki images are reused.

Deployment

GitHub Pages deployment runs on pushes to main and by manual dispatch. The workflow runs Python tests, data validation, a safe portrait dry run, and the Node.js test before uploading the unchanged static site. The deploy job runs only if all checks pass.

In the GitHub repository settings, set Pages to use GitHub Actions. Push to main, or open the Actions tab and run the Pages workflow manually. The workflow uses read-only repository access plus the Pages and identity-token permissions needed by the official Pages deploy action.

Sources

• A Wiki of Ice and Fire
• A Wiki of Ice and Fire copyright notice
• Creative Commons Attribution-ShareAlike 3.0
• Wikipedia List of Game of Thrones characters
• Creative Commons Attribution-ShareAlike 4.0
• An API of Ice and Fire documentation
• Game of Thrones Wiki character category
• Fandom licensing notice
• GitHub Pages custom workflows