Sergey Kostenchuk I build AI news infrastructure, Telegram automation, local knowledge workflows, and agentic developer tools. Current open-source focus: OpenCut OpenCode Copilot - OpenCut fork with OpenCode/Copilot action planning and…
Sergey Kostenchuk
I build AI news infrastructure, Telegram automation, local knowledge workflows, and agentic developer tools.
Current open-source focus:
- OpenCut OpenCode Copilot - OpenCut fork with OpenCode/Copilot action planning and semantic video editing workflows.
- MCP-QGIS - MCP/QGIS bridge for LLM-driven geospatial tool workflows.
- QGIS-OC-CHAT - QGIS chat plugin and OpenCode sidecar prototype for GIS tool workflows.
Application Package
The current Codex for Open Source application package is built around:
- a primary OpenCode/Copilot integration in a real open-source video editor;
- a supporting QGIS/GIS AI tooling ecosystem;
- reproducible cleanup, validation, release, and documentation workflows.
-
seo-llm-skill-cluster
Task-plan driven SEO and LLM-friendly website skill cluster
Python ★ 34 1mo agoExplain → -
task-plan-v2-dashboard
TASK-PLAN v2 dashboard and templates for auditable multi-agent software work
JavaScript ★ 31 17d agoExplain → -
45telega
Telegram MCP server with 45+ methods, QR auth, WSL/Kiro setup, and Docker-ready deployment
Python ★ 17 1mo agoExplain → -
MCP-QGIS
MCP/QGIS bridge for LLM-driven geospatial tool workflows
Python ★ 10 1mo agoExplain → -
Auto-iFlow-3.1.1-code
No description.
TypeScript ★ 9 2mo agoExplain → -
perplexity-pro-mcp
High-performance MCP server for Perplexity API integration with multi-threading support
★ 8 1y agoExplain → -
xray-xhttp-vpn-orchestrator
Sanitized public skill package for Xray/VLESS/XHTTP VPN orchestration
Python ★ 7 7d agoExplain → -
skill-creator-v2
Production-grade meta-skill for creating, evaluating, hardening, and packaging AI skills and skill groups
Python ★ 7 8d agoExplain → -
mlllm-threads-mcp
Enhanced Threads (Meta) MCP server with 50+ methods for Claude AI integration
TypeScript ★ 6 9mo agoExplain → -
ui-ux-agent-skill-system
Vendor-neutral UI/UX agent skill system with Codex, Claude, Gemini, Qwen, Copilot, GLM, Kimi adapters
Python ★ 4 23d agoExplain → -
ui-pattern-source-orchestrator
No description.
Python ★ 2 3d agoExplain → -
mouse-trail-masking-reveal
Mouse Trail Masking Reveal: portable Agent Skills for cursor-trail reveal effects across Codex, Claude Code, Gemini, Kimi, Qwen, and GLM agents
Python ★ 2 28d agoExplain → -
skill-portfolio-auditor
Evidence-first multi-provider skill portfolio auditor for Codex and compatible filesystem skill ecosystems
Python ★ 1 4d agoExplain → -
inferless-onboarding ⑂
Using Template for Model Deployment
★ 1 1y agoExplain → -
deepseek-r1-distill-qwen-32b ⑂
A distilled DeepSeek-R1 variant built on Qwen2.5-32B, fine-tuned with curated data for enhanced performance and efficiency. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>
★ 1 1y agoExplain → -
stable-video-diffusion ⑂
This model converts a single still image into a coherent video sequence with consistent, realistic motion. <metadata> gpu: T4 | collections: ["Diffusers"] </metadata>
★ 1 1y agoExplain → -
Wan2.1 ⑂
Wan: Open and Advanced Large-Scale Video Generative Models
Python ★ 1 1y agoExplain → -
threads-mcp ⑂
Threads (Meta) MCP server
Shell ★ 1 11mo agoExplain → -
afoc
AFOC (Auto Flow Open Code): Auto-iFlow experimental repo for OpenCode CLI integration
TypeScript ★ 1 2mo agoExplain → -
opencut-opencode-copilot
OpenCut fork with OpenCode/Copilot action planning and semantic video editing workflow
TypeScript ★ 1 1mo agoExplain → -
QGIS-OC-CHAT
QGIS chat plugin and OpenCode sidecar prototype for GIS tool workflows
Python ★ 1 1mo agoExplain → -
sergekostenchuk
GitHub profile README
★ 1 1mo agoExplain → -
codex-limit-bar
Lightweight native macOS menu bar utility for local Codex usage limits
Swift ★ 0 5d agoExplain → -
notebooklm-browser-research-skills
No description.
JavaScript ★ 0 10d agoExplain → -
antipenalty-legal-skills-system
Fail-closed legal skills system for ANTIPENALTY traffic fine defense workflows
JavaScript ★ 0 22d agoExplain → -
qwen3-14b ⑂
14B model with hybrid approach to problem-solving with two distinct modes: "thinking mode," which enables step-by-step reasoning and "non-thinking mode," designed for rapid, general-purpose responses. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>
★ 0 1y agoExplain →
No repos match these filters.