AI & ML

ivanshamaev Skill

Trino federated query architecture across heterogeneous sources — cross-catalog JOIN patterns (Iceberg+PostgreSQL+MySQL+Kafka+ClickHouse), pushdown behavior per connector type, minimizing cross-catalog data movement, materializing JDBC data into Iceberg, query routing strategy, connector-specific limitations (JDBC serial fetch, Kafka read-once), performance cost model for federated joins, metadata caching, CREATE TABLE AS SELECT federation patterns, catalog isolation design

ivanshamaev Skill

Ray Data distributed data processing — Dataset API, read_parquet/read_csv/read_json/read_delta_sharing, map/filter/flat_map/map_batches, groupby/aggregations, Actors for stateful transforms, Ray remote functions, write_parquet/write_iceberg, streaming execution, GPU batch inference, integration with Spark and Airflow

ivanshamaev Skill

OpenLineage data lineage tracking — RunEvent/Job/Dataset/facet spec, Marquez backend setup, Airflow/Spark/dbt integrations, column-level lineage, custom emitters, lineage-based impact analysis

ivanshamaev Skill

dbt + StarRocks models — dbt-starrocks adapter setup (profiles.yml), all materializations (table/view/incremental/ephemeral), incremental strategies (append/insert_overwrite/unique_key merge), StarRocks-specific model config (engine/keys/partition_by/distributed_by/properties), Duplicate/Aggregate/Primary Key table DDL from dbt, partition_by with date_trunc, on_schema_change behavior

ivanshamaev Skill

Workflow orchestration comparison and review — Airflow vs Prefect vs Dagster vs Temporal (architecture, execution model, dependency management, observability), when to choose each, migration patterns from Airflow to Prefect/Dagster, orchestrator evaluation scorecard, event-driven vs schedule-based triggering, Software-Defined Assets (Dagster) vs task-based (Airflow), hybrid orchestration patterns

asleekgeek Skill

Create tldr summaries for GitHub Copilot files (prompts, agents, instructions, collections), MCP servers, or documentation from URLs and queries.

asleekgeek Bundle

Convert any GitHub Copilot marketplace artefact (standalone agent, instruction fragment, plugin bundle, skill, hook, agentic workflow, cookbook recipe) into its Claude Code-native equivalent. Use this skill whenever the user wants to port, convert, adapt, or import an artefact from awesome-copilot or any Copilot-format plugin repo into Claude Code. Triggers on: "convert Copilot plugin", "port this agent to Claude Code", "import from awesome-copilot", "port the instructions", "convert this hook", "add this team to Claude Code", or when browsing the awesome-copilot marketplace and selecting an artefact to use. Also triggers for batch catalogue builds. Handles local repo sources and external GitHub repos. Produces ready-to-install agent .md files, adapted SKILL.md files, instruction files, hook bundles with claude-settings.json fragments, workflow prompts with GH Action templates, cookbook recipes, a per-plugin README with installation steps, and a running CATALOGUE.md entry.

asleekgeek Skill

Convert a text-based document to markdown following instructions from prompt, or if a documented option is passed, follow the instructions for that option.

asleekgeek Bundle

Phase execution skill for the Quality Playbook. Read by qp-orchestrator and passed to each phase sub-agent via the Task tool. Contains the seven-phase audit protocol, Mode A walkthrough instructions, domain exploration framework, role taxonomy, and challenge gate. Do not invoke this skill directly — invoke qp-orchestrator instead. Upstream: https://github.com/andrewstellman/quality-playbook

ivanshamaev Skill

StarRocks self-healing automation — auto-restart failed Routine Load jobs (detect PAUSED/CANCELLED + RESUME), rebalance tablet distribution after BE add/remove (REBALANCE command), trigger manual compaction for high-score tablets, auto-ANALYZE stale statistics, partition management (auto-drop old partitions), dead load label cleanup, BE disk space alerts + auto-tiering, scheduled health check + remediation agent

asleekgeek Skill

Bootstrap and run a multi-agent AI development team. Use when: starting a new software project with AI agents, setting up parallel dev/QA teams, creating sprint plans, writing brainstorm prompts with distinct agent voices, recovering a project workflow, or planning sprints. Each agent runs in a separate Claude Code session; the human acts as the message bus between sessions.

yiskang Bundle

Use this skill whenever you need to interact with the user's Anytype workspace through the anytype-mcp MCP server. Triggers whenever the user mentions Anytype, asks to create/find/update/delete notes, pages, tasks, or any objects in Anytype, wants to search their knowledge base, manage spaces or members, work with collections or lists, set up object types or properties, or use tags. Also triggers for GTD workflows (inbox, waiting, someday), task management in Anytype, or any question about what's in their Anytype space. If the user references their notes, their space, their tasks, or their wiki and you have the anytype-mcp tools available, use this skill.

anilveersingh1308 Bundle

Building LLM-Powered Applications with Claude

pangzhenying2025 Bundle

Python tool router, LLM council, adapters, detectors, KB builder. 73 Python files, 0 shell scripts.

pangzhenying2025 Bundle

39 agent definitions for specialized automotive roles: safety engineer, ADAS architect, AUTOSAR expert, BMS engineer, cybersecurity analyst, SOTIF specialist, and more.

cdesktop-ai Skill

Operate the cdesktop coding-session environment — agent teams, session management, file conventions.

anilveersingh1308 Bundle

Cross-model benchmark for gstack skills. Runs the same prompt through Claude, GPT (via Codex CLI), and Gemini side-by-side — compares latency, tokens, cost, and optionally quality via LLM judge. Answers "which model is actually best for this skill?" with data instead of vibes. Separate from /benchmark, which measures web page performance. Use when: "benchmark models", "compare models", "which model is best for X", "cross-model comparison", "model shootout". (gstack) Voice triggers (speech-to-text aliases): "compare models", "model shootout", "which model is best".

anilveersingh1308 Bundle

Pair a remote AI agent with your browser. One command generates a setup key and prints instructions the other agent can follow to connect. Works with OpenClaw, Hermes, Codex, Cursor, or any agent that can make HTTP requests. The remote agent gets its own tab with scoped access (read+write by default, admin on request). Use when asked to "pair agent", "connect agent", "share browser", "remote browser", "let another agent use my browser", or "give browser access". (gstack) Voice triggers (speech-to-text aliases): "pair agent", "connect agent", "share my browser", "remote browser access".

anilveersingh1308 Bundle

Keep gbrain current with this repo's code and refresh agent search guidance in CLAUDE.md. Wraps the gstack-gbrain-sync orchestrator with state probing, native code-surface registration, capability checks, and a verdict block. Re-runnable, idempotent. Use when: "sync gbrain", "refresh gbrain", "re-index this repo", "gbrain search isn't finding things". (gstack)

sunki1986 Bundle

Build, run, or adapt a multi-agent SOP assistant using LangGraph, LangChain, OpenAI, UV, and a JSON-backed policy knowledge base. Use when Codex needs to implement or maintain a single-file Python SOP assistant with a supervisor agent, intent classifier agent, SOP retrieval agent, answer generator agent, shared graph state, tool-based routing, and HR/Finance/IT policy question answering.

anilveersingh1308 Bundle

Set up gbrain for this coding agent: install the CLI, initialize a local PGLite or Supabase brain, register MCP, capture per-remote trust policy. One command from zero to "gbrain is running, and this agent can call it." Use when: "setup gbrain", "connect gbrain", "start gbrain", "install gbrain", "configure gbrain for this machine". (gstack)

houjingyi00417-hub Skill

Search scientific papers and retrieve structured experimental data extracted from full-text studies via the BGPT MCP server. Returns 25+ fields per paper including methods, results, sample sizes, quality scores, and conclusions. Use for literature reviews, evidence synthesis, and finding experimental details not available in abstracts alone.

houjingyi00417-hub Skill

Structured hypothesis formulation from observations. Use when you have experimental observations or data and need to formulate testable hypotheses with predictions, propose mechanisms, and design experiments to test them. Follows scientific method framework. For open-ended ideation use scientific-brainstorming; for automated LLM-driven hypothesis testing on datasets use hypogenic.

houjingyi00417-hub Skill

Bayesian modeling with PyMC. Build hierarchical models, MCMC (NUTS), variational inference, LOO/WAIC comparison, posterior checks, for probabilistic programming and inference.

houjingyi00417-hub Skill

Model interpretability and explainability using SHAP (SHapley Additive exPlanations). Use this skill when explaining machine learning model predictions, computing feature importance, generating SHAP plots (waterfall, beeswarm, bar, scatter, force, heatmap), debugging models, analyzing model bias or fairness, comparing models, or implementing explainable AI. Works with tree-based models (XGBoost, LightGBM, Random Forest), deep learning (TensorFlow, PyTorch), linear models, and any black-box model.

idoforgod Bundle

결심·계획이 *습관·일상 시스템*으로 정착하도록 설계하는 실행 지속력 코칭 스킬. BJ Fogg(Tiny Habits, 2019) + James Clear(Atomic Habits, 2018) + Charles Duhigg(The Power of Habit, 2012) 3대 습관 과학 토대를 통합. 5단계 습관 설계 — ① **목표를 *행동*으로 분해** (추상 → 구체 동작) ② **트리거(Cue) 설계** — 기존 습관에 *앵커링* (Habit Stacking) ③ **행동 작게 만들기(Tiny)** — 30초 안에 가능한 최소 단위 ④ **즉각 보상(Reward) 설계** — 도파민 보상으로 신경회로 강화 ⑤ **추적·축하(Tracking & Celebration)** — 가시화 + 작은 승리 자축. 한국 환경 특수성 반영 — 가족·교회·직장 시간 구조, *작심삼일* 패턴(시작 3일 후 동기 급락) 분석·대응. vision-readiness Follow-Through 점수 활용 — 점수 낮을수록 *더 작게·더 자주·더 가시적으로*. 사용자가 "습관 설계", "작심삼일", "실행이 안 돼요", "habit design", "Tiny Habits", "Atomic Habits", "꾸준함", "루틴 만들기", "B=MAP", "Habit Stacking", "Cue Routine Reward", "Never miss twice", "21일 신화", "Lally 66일"을 언급하거나 결심·계획이 행동으로 이어지지 않는 막막함을 호소할 때 발동한다. vision-strategy-coach의 Step 4(행동 계획·실행)를 *단독 깊이 도구*로 펼친 것. 박사님 강의 청중·교회 청년부·신년 결심자·다이어트·학습 습관자·박사님 본인 집필 루틴을 위해 설계되었다. 모든 사실 인용·날짜 계산·5단계 검증은 `habit_lib.py` 결정론 함수가 처리하며 LLM 자연어 추론을 차단한다.

houjingyi00417-hub Skill

RNA velocity analysis with scVelo. Estimate cell state transitions from unspliced/spliced mRNA dynamics, infer trajectory directions, compute latent time, and identify driver genes in single-cell RNA-seq data. Complements Scanpy/scVI-tools for trajectory inference.

houjingyi00417-hub Skill

Diffusion-based molecular docking. Predict protein-ligand binding poses from PDB/SMILES, confidence scores, virtual screening, for structure-based drug design. Not for affinity prediction.

duaooo Skill

Run read-only deep repository analysis and return a ranked synthesis with explicit confidence, concrete file references, and clear evidence-vs-inference boundaries. Use when a user says 'analyze', 'investigate', 'why does', 'what's causing', or needs grounded cross-file explanation before any changes are proposed.

idoforgod Bundle

사용자의 *꿈 달성 준비도(Dream Readiness)*를 20문항 자가 진단으로 평가하고 영어 축 막대 그래프로 시각화하는 비전 코칭 스킬. **결정론 환원** — 점수 파싱·범위/누락 검증·능력별 평균·Composite·레벨 매핑·강점/성장 분류·처방 스킬 매핑·ASCII 1/8 partial block 매핑·Mermaid 차트·재진단 델타는 *반드시* 동봉된 `lib/readiness_engine.py`(또는 `scripts/run_assessment.py`) 결정론 모듈을 호출해 산출한다 (LLM 자연어 추정 금지·할루시네이션 차단). 4가지 핵심 능력 — ① Seeing the Big Picture(큰 그림 보기) ② Reframing Inspiration into Realistic Goals(영감을 현실 목표로 재구성) ③ Creating Strategies to Achieve Goals(목표 달성 전략 수립) ④ Following Through on Plans(계획 실행 지속력) — 을 각 5문항씩 균형 있게 측정한다. 모든 20문항은 서로 다른 비유·다른 단어·다른 표현으로 작성되어 중복 없으며, 형식은 일관되게 "I am good at [skill]" 1인칭 진술이다. 응답은 0~10 척도(10=World Class, 7+ = 강점 영역, 4-6 = 평균 영역, 1-3 = 약점 영역, 0 = Not Yet 시작 전)이며, 십대 학생부터 어른까지 모두 즉시 이해 가능한 일상 비유(비디오 게임 테크 트리·마음의 언덕·Plan B·다음 주 금요일 등)로 표현된다. 20문항 응답 완료 후 막대 그래프(축·변수·라벨 모두 영어)로 4능력 점수 시각화하고, 강점·성장 영역·강화 방향을 한국어로 코칭한다. 사용자가 "비전 준비도", "꿈 달성 능력", "vision planning test", "readiness assessment", "20문항 비전 테스트", "내가 꿈을 이룰 능력이 있는가", "비전 코칭 시작", "큰 그림·실행력·전략 진단", "I am good at 테스트", "Big P

idoforgod Bundle

TLDR — Wild Cards Sub-skill ② Assessment. Petersen & Steinmüller(2009) V3.0 10장 Section III.2 "Assessment: Which are the most important Wild Cards for me or my organization?" 풀 구현 **INTERNAL** sub-skill. Petersen 4-Factor Pyramid (Being·Sustenance·Actions·Tools) + Power Factor 1-4 적용 (Being=4 highest, Tools=1 lowest). 4 Major Category × 변수 풀 평가 — Being(perception of reality·strongly held personal values·health/wellness·physical environment) · Sustenance(location/habitat·food and water·energy·transport) · Actions(personal relations·formal group relations·work and recreation) · Tools(communicate·learn·make/distribute things technology). Process of elimination (PDF p.5 "relative process · biases consistent"). Target group filter (PDF p.3 "close to home" variation). Identification output 20-40 candidates → top N filter (default 10, valid: {5,10,15,20}). 외부 사람·평가단 미동원 — AI Petersen Pyramid Evaluator Agent 자동 작동. **결정론 강제**: wc_assessment_validator.py 호출 — PPS 계산·Affinity 정규화·Top-N 필터·다양성 검증 LLM 재추론 금지. 4 Major Cat

houjingyi00417-hub Skill

Observe the user's screen via screenpipe, detect repeated research workflows, match them against existing scientific-agent-skills, and draft new skills (or composition recipes that chain existing ones) for the patterns not yet covered. Use when the user asks to analyze their recent work and propose skills based on what they actually do. Requires the screenpipe daemon (https://github.com/screenpipe/screenpipe) running locally on port 3030 鈥?the skill has no other data source and will refuse to run if screenpipe is unreachable. All detection runs locally; only redacted cluster summaries reach the LLM.

houjingyi00417-hub Skill

Automated LLM-driven hypothesis generation and testing on tabular datasets. Use when you want to systematically explore hypotheses about patterns in empirical data (e.g., deception detection, content analysis). Combines literature insights with data-driven hypothesis testing. For manual hypothesis formulation use hypothesis-generation; for creative ideation use scientific-brainstorming.

houjingyi00417-hub Skill

High-performance reinforcement learning framework optimized for speed and scale. Use when you need fast parallel training, vectorized environments, multi-agent systems, or integration with game environments (Atari, Procgen, NetHack). Achieves 2-10x speedups over standard implementations. For quick prototyping or standard algorithm implementations with extensive documentation, use stable-baselines3 instead.

houjingyi00417-hub Skill

Chat with your agent about projects, recommendations, and canonical papers in Paperzilla. Use when users ask for recent project recommendations, canonical paper details, markdown-based summaries, recommendation feedback, feed export, or Atom feed URLs.

duaooo Bundle

This skill should be used when the user asks to "analyze experimental results", "generate results section", "statistical analysis of experiments", "compare model performance", "create results visualization", or mentions connecting experimental data to paper writing. Provides comprehensive guidance for analyzing ML/AI experimental results and generating paper-ready content.