Browse Skills

fine-tuning-with-trl.md

62

1 export fine-tuning-with-trl

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.

2026-01-16

mcp-builder.md

62

1 export mcp-builder

2 from "xenitV1/claude-code-maestro"

3 MCP (Model Context Protocol) server building principles. Tool design, resource patterns, best practices.

2026-01-15

llava.md

62

1 export llava

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models. Supports multi-turn image chat, visual question answering, and instruction following. Use for vision-language chatbots or image understanding tasks. Best for conversational image analysis.

2026-01-16

testing-skills-with-subagents.md

62

1 export testing-skills-with-subagents

2 from "mneves75/dnschat"

3 Use when creating or editing skills, before deployment, to verify they work under pressure and resist rationalization - applies RED-GREEN-REFACTOR cycle to process documentation by running baseline without skill, writing to address failures, iterating to close loopholes

2026-01-13

onboarding.md

62

1 export onboarding

2 from "huytieu/COG-second-brain"

3 Personalize COG for your workflow - creates profile, interests, and watchlist files with guided setup (run this first!)

2026-01-15

sglang.md

62

1 export sglang

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, constrained decoding, agentic workflows with tool calls, or when you need 5× faster inference than vLLM with prefix sharing. Powers 300,000+ GPUs at xAI, AMD, NVIDIA, and LinkedIn.

2026-01-16

pc-games.md

62

1 export pc-games

2 from "xenitV1/claude-code-maestro"

3 PC and console game development principles. Engine selection, platform features, optimization strategies.

2026-01-15

conversation-manager.md

62

1 export conversation-manager

2 from "xenitV1/claude-code-maestro"

3 Smart dialogue manager for user communication. Clarifies unclear requests, asks the right questions, provides progress updates. Use when starting new projects, handling unclear requests, or communicating build status.

2026-01-15

llama-factory.md

62

1 export llama-factory

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support

2026-01-16

constitutional-ai.md

62

1 export constitutional-ai

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with self-critique/revision, then RLAIF (RL from AI Feedback). Use for safety alignment, reducing harmful outputs without human labels. Powers Claude's safety system.

2026-01-16

knowledge-consolidation.md

62

1 export knowledge-consolidation

2 from "huytieu/COG-second-brain"

3 Build frameworks from scattered insights across all braindumps and notes

2026-01-15

vulnerability-scanner.md

62

1 export vulnerability-scanner

2 from "xenitV1/claude-code-maestro"

3 Advanced vulnerability analysis principles. OWASP 2025, Supply Chain Security, attack surface mapping, risk prioritization.

2026-01-15

app-builder.md

62

1 export app-builder

2 from "xenitV1/claude-code-maestro"

3 Main application building orchestrator. Creates full-stack applications from natural language requests. Determines project type, selects tech stack, coordinates agents.

2026-01-15

react-patterns.md

62

1 export react-patterns

2 from "xenitV1/claude-code-maestro"

3 Modern React patterns and principles. Hooks, composition, performance, TypeScript best practices.

2026-01-15

plan-writing.md

62

1 export plan-writing

2 from "xenitV1/claude-code-maestro"

3 Structured task planning with clear breakdowns, dependencies, and verification criteria. Use when implementing features, refactoring, or any multi-step work.

2026-01-15

brainstorming.md

62

1 export brainstorming

2 from "mneves75/dnschat"

3 Use when creating or developing anything, before writing code or implementation plans - refines rough ideas into fully-formed designs through structured Socratic questioning, alternative exploration, and incremental validation

2026-01-13

dispatching-parallel-agents.md

62

1 export dispatching-parallel-agents

2 from "mneves75/dnschat"

3 Use when facing 3+ independent failures that can be investigated without shared state or dependencies - dispatches multiple Claude agents to investigate and fix independent problems concurrently

2026-01-13

systematic-debugging.md

62

1 export systematic-debugging

2 from "mneves75/dnschat"

3 Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes - four-phase framework (root cause investigation, pattern analysis, hypothesis testing, implementation) that ensures understanding before attempting solutions

2026-01-13

database-design.md

62

1 export database-design

2 from "xenitV1/claude-code-maestro"

3 Database design principles and decision-making. Schema design, indexing strategy, ORM selection, serverless databases.

2026-01-15

testing-anti-patterns.md

62

1 export testing-anti-patterns

2 from "mneves75/dnschat"

3 Use when writing or changing tests, adding mocks, or tempted to add test-only methods to production code - prevents testing mock behavior, production pollution with test-only methods, and mocking without understanding dependencies

2026-01-13

braindump.md

62

1 export braindump

2 from "huytieu/COG-second-brain"

3 Quick capture of raw thoughts with intelligent domain classification and competitive intelligence extraction

2026-01-15

evaluating-llms-harness.md

62

1 export evaluating-llms-harness

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

2026-01-16

optimizing-attention-flash.md

62

1 export optimizing-attention-flash

2 from "zechenzhangAGI/AI-research-SKILLs"

3 Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

2026-01-16

seo-fundamentals.md

62

1 export seo-fundamentals

2 from "xenitV1/claude-code-maestro"

3 SEO fundamentals, E-E-A-T, Core Web Vitals, and Google algorithm principles.

2026-01-15