Tools¶
GameForge ships fourteen AI tools. With USE_MOCK_AI=true they run without paid keys; with USE_MOCK_AI=false they call external providers (OpenAI / ProxyAPI and optional services).
Marketing pages: /tools/<slug> (EN) and /ru/tools/<slug> (RU). In-app editors: https://gameforge.website/<slug> after login.
Level Designer¶
- Input: natural-language level brief (optional project)
- Output: tilemap JSON + canvas preview
- API:
POST /api/v1/level-designer
Quest Generator¶
- Input: quest / story prompt
- Output: quest structure, dialogues, branches as JSON
- API:
POST /api/v1/quest-generator
Texture Upscaler¶
- Input: uploaded image + scale (2× / 4×)
- Output: upscaled PNG URL (Real-ESRGAN service or PIL mock)
- API:
POST /api/v1/texture-upscaler - Heavy work may run on the Celery worker
Character Creator¶
- Input: character description (required; regenerate keeps last body)
- Output: image URL in object storage
- API:
POST /api/v1/character-creator
Sound Designer¶
- Input: prompt + kind (sfx / music / voice)
- Output: audio file URL (mock waveform or provider)
- API:
POST /api/v1/sound-designer
Playtester¶
- Input: design doc / feature description
- Output: QA-style report JSON (issues, severity, suggestions)
- API:
POST /api/v1/playtester
Localization¶
- Input: source strings / JSON + target languages
- Output: translated JSON or CSV
- API:
POST /api/v1/localization
Game Balancer¶
- Input: game data (classes, weapons, economy stats as JSON)
- Output: balance metrics, outliers, and tuning suggestions
- API:
POST /api/v1/game-balancer
Level Analyzer¶
- Input: level / path data for difficulty and reachability analysis
- Output: pathfinding summary, difficulty signals, choke points
- API:
POST /api/v1/level-analyzer - Compare:
POST /api/v1/level-analyzer/compare— side-by-side analysis of two levels
Store Description¶
- Input: game brief, platform (Steam / App Store / Google Play), tone, language
- Output: store listing copy (title, short/long description, tags, keywords)
- API:
POST /api/v1/store-description - UI language (
gf_lang) influences default output language when the UI is Russian
Playtest Analyzer¶
- Input: playtest session metrics / feedback notes
- Output: retention insights, friction points, prioritized recommendations
- API:
POST /api/v1/playtest-analyzer
Playtester vs Playtest Analyzer
Playtester reviews a design document before build. Playtest Analyzer interprets telemetry and feedback after sessions.
Trailer Script¶
- Input: game pitch, trailer type (launch / teaser / gameplay), duration, tone
- Output: timed scenes, voiceover, text overlays, sound-design notes
- API:
POST /api/v1/trailer-script
Review Analyzer¶
- Input: player reviews (text, optional ratings / language)
- Output: sentiment, recurring issues, theme clusters, average rating signals
- API:
POST /api/v1/review-analyzer
Discord Bot (studio)¶
Configure a community bot (token encrypted at rest), commands, moderation helpers, and analytics. This is a multi-endpoint tool — not a single POST.
| Action | Method | Path |
|---|---|---|
| Save / update config | POST |
/api/v1/discord-bot/configure |
| Status | GET |
/api/v1/discord-bot/status |
| List commands | GET |
/api/v1/discord-bot/commands |
| Create command | POST |
/api/v1/discord-bot/command |
| Moderate message | POST |
/api/v1/discord-bot/moderate |
| Simulate command | POST |
/api/v1/discord-bot/simulate-command |
| Analyze community | POST |
/api/v1/discord-bot/analyze |
| Analytics snapshot | GET |
/api/v1/discord-bot/analytics |
| Tracked users | GET |
/api/v1/discord-bot/users |
There is no live Discord gateway worker in the default deploy — moderation and commands are API/simulate MVP. Tokens are Fernet-encrypted before storage.
Asset URLs¶
Generated files are stored in MinIO. In production, browsers open signed URLs under https://<domain>/s3/.... If a link fails, check Deployment → VPS (S3_PUBLIC_ENDPOINT).