farkhad
Community@f4rkh4d · Almaty, Kazakhstan
software engineer · almaty
Agent Skills by farkhad
Showing 51 vetted skills indexed across 1 GitHub repositories.
forge-research
Convert broad research requests into sourced, confidence-scored claims.
forge-citation
Enforces citationable claims and validates URLs/files in AI-generated documents.
forge-web-search
Guide AI agents through credible web research with bounded fetching and citation verification.
forge-brutalist
Generate brutalist UI layouts with strict grids, mono typography, and one accent color.
forge-react-hooks
Detect React hook order violations using AST-based analysis.
forge-frontend-nextjs
Enforce Next.js 16 app-router server-first patterns in React code.
forge-frontend
Enforce frontend UI style constraints for React/TSX generation.
forge-accessibility
Enforce semantic HTML, keyboard navigation, and WCAG contrast in frontend UI.
forge-soft
Apply premium-soft UI styling with warm neutrals, refined typography, and unified shadows.
forge-performance
Enforce Core Web Vitals budgets for React, Next.js, and Vite interfaces.
forge-redesign
Audit existing web apps and apply surgical Tailwind styling improvements.
forge-minimalist
Enforce editorial minimalist UI rules for three-color palettes and typography.
forge-observability
Standardize production observability with correlated traces, metrics, and logs.
forge-caddy
Harden Caddyfile configurations for production traffic with safe defaults.
forge-dockerfile
Generate and validate secure, deterministic Dockerfiles for CI builds.
forge-kubernetes
Generate Kubernetes manifests with pinned images, probes and hardened security settings.
forge-github-actions
Enforce security and reliability hygiene rules for GitHub Actions workflows.
forge-llm-streaming
Stream LLM responses to browser UIs via SSE with typed events and cancellation.
forge-tool-use
Enforce capped, validated tool-use loops with structured error handling.
forge-rag
Design retrieval-augmented generation pipelines with hybrid retrieval, reranking, citations, and refusal behavior.
forge-evals
Design, score, and run evals for LLM features with golden datasets and rubrics.
forge-prompt-engineering
Audit production LLM prompts for structured outputs and injection defenses.
forge-secrets
Enforce safe secret loading, logging, error handling, storage, and rotation.
forge-full-output
Enforce complete file output and ban truncation placeholders in generated code.