Claude Code hook that enforces safe file editing practices. Prevents data loss during bulk changes.

Hunspell-based Turkish diacritics validator. Detects missing Turkish characters in content files.

Local semantic code search with Tree-sitter AST chunking. LanceDB + sentence-transformers hybrid search.

Knowledge management MCP server for markdown and code files. FTS5 + sqlite-vec hybrid search.

My own ComfyUI node package. It analyses a texture's vein topology and splits NxM tiling along the structure, solving the broken-pattern problem of a plain grid cut.

My own ComfyUI node. I use it to merge texture patches without leaving visible seams.

Style and mask LoRAs I trained on RunPod. I use them to carry my own visual language into the ComfyUI pipeline.

I use it on a subscription for everyday work, research and NotebookLM. I rarely reach for it when coding.

Moonshot AI's long-context model. I use it on a subscription for long texts, document analysis, bug fixes and reporting.

I use it for coding and planning, mostly through Claude Code in the terminal.

OpenAI's coding agent. I use it on a subscription for second opinions and parallel work.

Cloud infrastructure platform. I use it to deploy applications quickly and easily.

Platform providing CDN, security, and infrastructure services.

Germany-based cloud and dedicated server provider. For affordable and reliable infrastructure.

Serverless PostgreSQL database. With branching and connection pooling features.

Event-driven background job platform. For webhook processing and async task management.

GPU cloud platform. I use it for LoRA model training and inference.

Open source self-hosted PaaS. For deploying and managing applications on Hetzner.

Google's AI notebook for note-taking and research.

Google Cloud's enterprise AI platform. For scalable AI solutions.

CLI tool for running local LLM models.

Data science and machine learning platform. Notebooks, datasets, and competitions.

Open source image generation model. Run on your own server.

Node-based Stable Diffusion interface. For LoRA, ControlNet, and complex workflows.

Black Forest Labs' open-weight image model. I run the Klein variant locally through ComfyUI.

Image generation and editing model. The largest model in my local ComfyUI pipeline, used for identity-preserving edits.

Open source raster editor and its plugin that connects to a local ComfyUI. For selective inpaint, outpaint and live sketching.

Google's image generation models (Imagen, ImageGen). High-quality image generation.

I use it for image and video processing models.

Realistic AI voice synthesis. Multi-language and voice cloning support.

Edit video/podcast like text. Transcription and AI features.

Alibaba's open source multilingual model family. I delegate coding work and everyday agentic tasks to it locally.

Google's open source model family. The model I run locally for Turkish and English content work.

Multilingual embedding model. It powers the vector layer of the semantic search in my own MCP servers.

Lightweight embedding model. I use it for local search work that needs fast indexing.

AI-powered automation. No-code workflows and integrations.

No-code automation platform. Workflow automation and integrations.

Open source workflow automation. Recently added AI features.

Database and automation platform. Enhanced with AI features.

Academic paper discovery and organization. Citation network analysis.

AI research assistant. Literature review and synthesis.

Google's AI-powered design tool. I use it to create initial draft ideas. Generates responsive UI designs from text prompts or image sketches.

Web-based design and prototyping tool. I handle the main design process with Figma. For UI/UX design and collaboration.