findinfinitelabs
- 22 skills
- 0 followers
- 6 hours ago last updated
- ▌ Bible Epub Processing · findinfinitelabsParse JW.org NWT EPUBs to extract individual Bible verses for verse previews, parallel-corpus building, and Bible-coverage analytics. Use when working with NWT EPUB files, building parallel Chuukese↔English Bible data, or wiring scripture preview features.
- ▌ CSS Styling Standards · findinfinitelabsStyling conventions for the Chuuk Dictionary frontend — Mantine v8 theme, CSS Modules per page, global app-shell CSS, multilingual / accented-character considerations. Use when adding or modifying styles in `frontend/src/`.
- ▌ Flask API Development · findinfinitelabsPatterns for the Chuuk Dictionary Flask backend — single-file `app.py` monolith, role-permission gated routes, Cosmos DB / `DictionaryDB` integration, SSE processing streams, and the translate/correction loop. Use when adding/altering API endpoints, wiring auth into new routes, or modifying request/response shapes.
- ▌ LLM Ollama Integration · findinfinitelabsHow the Chuuk Dictionary integrates Ollama for the Chuukese-translator custom LLM — `ChuukeseLLMTrainer`, the `chuukese-translator` modelfile, container/runtime wiring, and the translate endpoint integration. Use when changing LLM behavior, updating the modelfile, debugging Ollama connectivity, or modifying the LLM training pipeline.
- ▌ Python Venv Management · findinfinitelabsAutomatically manage Python virtual environments (.venv) in terminal commands. Always activate .venv before running Python/pip commands. Supports macOS, Linux, and Windows with shell-aware activation. Use when executing Python scripts, installing packages, or running development servers. Critical for consistent environment management.
- ▌ Docker Containerization · findinfinitelabsReality of the Chuuk Dictionary container builds — multi-stage Flask + React app image and a separate Ollama sidecar image. No docker-compose is used. Use when modifying the Dockerfiles, debugging build failures, or adding system dependencies.
- ▌ Document Ocr Processing · findinfinitelabsOCR pipeline for the Chuuk Dictionary — Tesseract + Google Vision via `OCRProcessor`, large-document handling via `EnhancedOCRProcessor`, structure-aware parsing via `AdvancedDocumentParser`. Use when ingesting scanned dictionary pages, PDFs, or DOCX files, or debugging extraction quality.
- ▌ Git Workflow Management · findinfinitelabsAutomate frequent Git operations including adding changes, committing with meaningful messages, and pushing to remote repositories. Use when you need to save progress, commit code changes, or maintain regular version control workflow with proper commit practices.
- ▌ Intelligent Text Chunking · findinfinitelabsSplit long Chuukese/English documents into chunks for OCR pipelines, AI training, or retrieval — using the in-repo `IntelligentTextChunker`. Use when a downstream consumer requires bounded context windows or when feeding large documents into the training generator.
- ▌ Large Document Processing · findinfinitelabsProcess large documents (200+ pages) with structure preservation, intelligent parsing, and memory-efficient handling. Also covers intelligent text chunking for AI training and RAG systems. Use when working with complex formatted documents, multi-level hierarchies, or when splitting large content for AI pipelines.
- ▌ React Typescript Frontend · findinfinitelabsPatterns for the Chuuk Dictionary React + TypeScript frontend — Vite, Mantine v8, React 19, React Router v7, axios with cookie sessions, and permission-gated routing. Use when adding pages, wiring API calls, modifying navigation, or extending the user/auth context.
- ▌ Azure Container Deployment · findinfinitelabsDeploy the Chuuk Dictionary stack (main app + Ollama sidecar) to Azure Container Apps. Covers ACR remote builds via `az acr build`, Key Vault prerequisites, Cosmos DB credential injection, and the env-var contract. Use when running a deploy, debugging a failed deploy, or modifying Azure infrastructure.
- ▌ AI Training Data Generation · findinfinitelabsGenerate high-quality training datasets from documents, text corpora, EPUBs, and structured content. Use when creating AI training data from dictionaries, Bible EPUBs, brochures, or when generating examples for machine learning models. Optimized for low-resource languages and domain-specific knowledge extraction. Supports parallel corpus extraction from NWT Bible EPUBs.
- ▌ Helsinki Nlp Model Training · findinfinitelabsFine-tuning the Helsinki-NLP OPUS-MT models for Chuukese ↔ English in this repo — `HelsinkiFineTuner` device selection, training-data assembly from `DictionaryDB`, BLEU evaluation, and where the trained models live. Use when modifying the trainer, adding evaluation metrics, debugging a training run, or changing model paths.
- ▌ Chuukese Language Processing · findinfinitelabsSpecialized processing for Chuukese language text including tokenization, accent handling, cultural context preservation, and language-specific patterns. Use when working with Chuukese text, translation tasks, or when building language models for this Micronesian language.
- ▌ Code Documentation Standards · findinfinitelabsComprehensive code documentation standards and guidelines for maintaining up-to-date documentation across Python, HTML, CSS, and JavaScript codebases. Use when creating or modifying code to ensure proper documentation practices and maintainable code.
- ▌
- ▌ Testing And Quality Assurance · findinfinitelabsPytest patterns for the Chuuk Dictionary backend — running the suite, the shared `mock_db` / `client` / `auth_headers` fixtures, the right places to patch (`DictionaryDB` class, not `app.dict_db`), and how to mark unit vs integration vs translation tests. Use when adding tests, fixing fixture failures, or wiring CI.
- ▌ Database Management Operations · findinfinitelabsConventions for the Chuuk Dictionary persistence layer — Azure Cosmos DB (MongoDB API) via `db_factory`, `DictionaryDB`, `UserDB`, and `PublicationManager`. Covers connection-resolution order, the actual collection/method names in use, and the managed-identity path. Use when adding queries, debugging connection issues, or extending the schema.
- ▌ Security Environment Standards · findinfinitelabsSecurity model and environment-variable conventions for the Chuuk Dictionary app — hybrid access-code + magic-link auth, role/permission gating, single-active-session enforcement, file-upload allow-list, scripture endpoint anti-scrape, and the env vars actually read. Use when adding secrets, configuring environments, or hardening a route.
- ▌ Translation Quality Assessment · findinfinitelabsAssess and validate translation quality between Chuukese and English with cultural context awareness, linguistic accuracy checking, and automated quality metrics. Supports Helsinki-NLP model evaluation with BLEU, chrF, and model-specific scoring. Use when evaluating translation outputs, building quality control systems, or validating translation models.
- ▌ Multi Language Document Processing · findinfinitelabsPatterns for documents that mix Chuukese + English (and occasionally other JW.org locales) — sentence segmentation, scripture-safe chunking, and per-language sentence stores. Use when ingesting bilingual brochures/articles or building parallel corpora.