Data and AI skills for Claude Code.

Defines custom DataPoint node and edge models so cognee's LLM extraction builds a typed knowledge graph.

33k stars

cognee-custom-pipelines

Makes the agent write custom cognee tasks, chain them into pipelines, and store custom DataPoints.

33k stars

cognee-forget

Makes the agent find and safely delete specific documents, datasets, or derived memory in cognee.

33k stars

cognee-improve-sessions

Makes the agent use cognee session memory and improve() to store turns and bridge them into the graph.

33k stars

cognee-ingestion

Makes the agent ingest text, files, URLs and repos into cognee with remember(), choosing extractors.

33k stars

cognee-install

Makes the agent install the cognee Python SDK and run a first remember and recall example.

33k stars

cognee-performance

Tunes cognee ingestion and reads for speed, rate limits and cost: batching, chunk size, dry-run estimates.

33k stars

cognee-permissions

Sets up cognee users, roles and dataset sharing, and explains how permissions and isolation are enforced.

33k stars

cognee-recall

Queries cognee memory with recall(), choosing search types, dataset scope and reading results or citations.

33k stars

company-brain-company-qa

muratcankoylan/Agent-Skills-for-Context-Engineering

Loads a company's SQL database, tickets and docs into cognee and answers questions across all three.

33k stars

add-atomic-chat-tool

nanocoai/nanoclaw

Lets the container agent call local models served by the Atomic Chat desktop app through MCP tools.

31k stars

agent-platform-eval-flywheel

google/skills

Guides evaluation of agents and models with datasets, LLM-as-judge metrics and failure analysis.

22k stars

advanced-evaluation

Teaches LLM-as-judge methods: direct scoring, pairwise comparison, rubric calibration and bias control.

18k stars

project-development

muratcankoylan/Agent-Skills-for-Context-Engineering

Guides project-level LLM decisions: task fit, pipeline shape, cost estimates, and single versus multi-agent choice.

18k stars

prompt-master

nidhinjs/prompt-master

Writes one optimized, paste-ready prompt for a named AI tool, such as an LLM, Cursor or Midjourney.

15k stars

chunking-embeddings

Guides changes to Xberg's text chunking config, chunker types and embedding generation for RAG pipelines.

9k stars

xberg-chunking

deanpeters/Product-Manager-Skills

Splits extracted text into chunks for LLM context or RAG, sized by characters or tokens, with overlap.

9k stars

antinet-four-color-cards

anbeime/skill

Turns a parsed document into sourced fact, explanation, risk and action cards for review.

9k stars

expert-prompt-migrator

anbeime/skill

Rewrites expert prompts into a portable five-part format for reuse across different AI assistants.

9k stars

yao-decision

YaoApp/yao

Gets typed classifications, scores, and calibrated probabilities from a decision model via the tai CLI.

8k stars

context-engineering-advisor

Diagnoses whether an AI workflow suffers from context stuffing and guides structured context design fixes.

8k stars

0g-compute

internet-court/internet-court-skill

Guides 0G Compute SDK calls for decentralized AI inference, fine-tuning and response handling.

7k stars

n8n-agents

czlonkowski/n8n-skills

Guides design of n8n AI agent, chain and classifier nodes covering tools, memory, output parsing and RAG.

7k stars

sn-image-base

Exposes SenseNova image generation, editing, VLM recognition and LLM text tools as raw backend calls.

6k stars

agentica-claude-proxy

Guides wiring the Agentica SDK through a local proxy that calls the Claude CLI with tool permissions.

4k stars

agentica-infrastructure

Reference for Agentica's multi-agent pattern classes, core components and coordination APIs.

4k stars

agentica-prompts

Explains how to write prompts for Agentica REPL agents so they follow instructions reliably.

4k stars

agentica-sdk

Shows how to build Python agents with the Agentica SDK using agentic functions, spawn and tools.

4k stars

agentica-server

Sets up the Agentica server with a Claude proxy and explains the startup order and debugging steps.

4k stars

agentica-spawn

atomicstrata/llm-wiki-compiler

Picks and launches Agentica multi-agent patterns such as swarm, hierarchical, generator-critic and jury.

4k stars

pinme-llm

glitternetwork/pinme

Makes the agent write PinMe Worker code that calls OpenRouter models and chat completions via a proxy.

4k stars

llm-provider-integration

samugit83/redamon

Adds an LLM provider to RedAmon while keeping API keys out of scan containers and model ids prefix-routed.

3k stars

llmwiki

Builds and queries a citation-traceable knowledge wiki from documents to retrieve evidence for tasks.

2k stars

fireworks-inference

Leanmcp/gateway-skills

Guides Fireworks AI inference: endpoints, keys, model ids, reasoning output quirks and tool calling.

2k stars

mcp-apps-builder

Leanmcp/gateway-skills

Builds MCP servers with interactive React UIs using the Leanmcp UI SDK decorators and components.

2k stars

patterns

pchalasani/claude-code-tools

Gives the agent an indexed list of Langroid multi-agent design patterns with reference docs to consult.

2k stars

create-mcp-servers

glittercowboy/taches-cc-resources

Guides the agent in building MCP servers that expose tools, resources and prompts to Claude.

2k stars

create-meta-prompts

glittercowboy/taches-cc-resources

Creates structured prompts for multi-stage Claude pipelines with research, planning and execution stages.

2k stars

build-mcp

NeoLabHQ/context-engineering-kit

Guides building MCP servers in Python FastMCP or TypeScript with well-designed, agent-friendly tools.

2k stars

cross-video-qa

JimLiu/baocut

Answers questions across a video library by semantic search, citing each claim with its video and timestamp.

2k stars

ai-evals

RefoundAI/lenny-skills

Guides building evaluation systems for AI products using human, code-based and LLM-as-judge methods.

1k stars

llm-provider

caliber-ai-org/ai-setup

Adds an LLM provider with call and stream methods, wired into the provider factory and usage tracking.

1k stars

apple-on-device-ai

dpearson2699/swift-ios-skills

Helps choose and deploy on-device AI on Apple devices with Foundation Models, Core ML, MLX Swift or llama.cpp.

1k stars

axiom-ai

CharlesWiltgen/Axiom

Guides building on-device Apple Intelligence features with Foundation Models, structured output, tools and evals.

1k stars

bedrock

itsmostafa/aws-agent-skills

Guides the agent through calling AWS Bedrock foundation models, creating embeddings and building RAG.

1k stars

mcp-maintainer

hoangsonww/Claude-Code-Agent-Monitor

Maintains this repo's local MCP server, covering tool domains, safety guards, host config and runtime troubleshooting.

1k stars

prompt-optimization

asklokesh/loki-mode

Repeats prompts for Haiku and other non-reasoning models to improve accuracy on structured tasks like tests and parsing.

1k stars

launch-your-agent

anthropics/launch-your-agent

Guides a founder to scope, launch, grade and schedule a Claude Managed Agent in their own account.

1k stars

wrap-up

anthropics/launch-your-agent

Recaps a finished Claude Managed Agent build, shows its resources and status, and suggests next steps.

1k stars

extract-atomic-memory

Extracts each independently changeable claim from a conversation window as a record with subject, time and status.

1k stars

judge-group-closure

Judges which offered memory rows a nominated statement retires, by meaning rather than shared words.

1k stars

judge-rem-memory-updates

Judges which older memory facts a newer record supersedes and rewrites the facts that remain current.

1k stars

judge-retired-position

Judges whether a stored row states a retired profile position as current, by meaning.

1k stars

judge-task-lifecycle

Judges whether a memory completes, removes or opens an active task, by meaning rather than wording.

1k stars

key-state-attribute

Makes the agent pick one offered state attribute slug for a stored statement, or null when none fits.

1k stars

resolve-entity-identity

GarethManning/education-agent-skills

Makes the agent match a new display name to an existing entity id, or answer new when unsure.

1k stars

ai-feedback-design-principles

Audits and redesigns AI-generated feedback for pedagogical quality, timing and learning impact.

929 stars

council-execution

hex/claude-council

Runs a question across several AI providers, shows each answer verbatim, and writes a consensus synthesis.

913 stars

deep-execution

hex/claude-council

Runs parallel Claude analyst agents across providers and returns schema-checked analyses with confidence ratings.

913 stars

local-council-execution

hex/claude-council

Convenes independent Claude subagents, each with one lens, and synthesizes their views as a same-model panel.

913 stars

ai-product

coco-research/coco

Guides production LLM features: validated outputs, RAG, guardrails, streaming and cost control.

795 stars

agentic-development

alinaqi/maggy

Guides building tool-using AI agents with Pydantic AI in Python and the Claude Agent SDK in Node.js.

710 stars

ai-models

alinaqi/maggy

Reference of current AI model names, specs and picking tips for Claude, OpenAI, Gemini and others.

710 stars

convex-agents

waynesutton/builder-skills

Builds chat and tool-calling agents on the Convex agent component with threads, streaming, and RAG.

405 stars

agent-eval-harness

OneWave-AI/claude-skills

Builds a small eval set with checkable pass criteria to test whether an agent, prompt or skill works and catches regressions.

400 stars

AgentDB Advanced Features

spencermarx/open-code-review

Covers advanced AgentDB use: QUIC sync, multiple databases, custom distance metrics and hybrid vector search.

382 stars

AgentDB Memory Patterns

spencermarx/open-code-review

Implements persistent agent memory with AgentDB for session history, long-term storage and context management.

382 stars

chunking-strategy

giuseppe-trisciuoglio/developer-kit

Recommends chunk sizes, overlap and boundary methods for RAG and checks retrieval precision and recall.

366 stars

prompt-engineering

giuseppe-trisciuoglio/developer-kit

Guides writing and optimizing LLM prompts with few-shot examples, chain-of-thought and system prompt design.

366 stars

rag

giuseppe-trisciuoglio/developer-kit

Guides building RAG pipelines with document chunking, embeddings, vector storage and retrieval.

366 stars

agent-interface-design

Neeeophytee/finding-unknowns-skills

Guides designing agent-facing tools and CLIs so parameters and descriptions teach correct use.

358 stars

prompt-architect

ckelsoe/prompt-architect

Scores and rewrites a prompt using 31 named frameworks, asking clarifying questions when intent is unclear.

327 stars

jeecg-aiflow

jeecgboot/skills

Manages JeecgBoot AIFlow LLM flows from natural language: create, edit, query, debug, publish, delete.

245 stars

freeride

Shaivpidadi/FreeRide

Ranks free OpenRouter models and sets fallbacks in openclaw.json to handle rate limits.

238 stars

ai-output-validation

DevelopersGlobal/ai-agent-skills

Enforces schema-defined structured outputs, validation, and fallbacks at AI pipeline boundaries.

216 stars

iris-development

redis/agent-skills

Guides the agent to record session events and search long-term memories using the Redis Agent Memory SDKs.

171 stars

chain-of-verification

serpro69/claude-toolbox

Has the agent check its own answers by generating and answering verification questions before revising.

154 stars

building-with-typesafe-jev

aaddrick/building-with-typesafe-jev

Guides using the Jev model for narrow typed judgments on natural language, such as classify, route, or score.

148 stars

route

ypollak2/llm-router

Classifies a task by type and complexity and routes it to the most suitable LLM.

120 stars

neo4j-agent-memory-skill

neo4j-contrib/neo4j-skills

Gives the agent reference guidance for the neo4j-agent-memory package and its hosted memory service.

118 stars

neo4j-aura-agent-skill

neo4j-contrib/neo4j-skills

Creates, configures, and invokes Neo4j Aura agents through the v2beta1 REST API.

118 stars

agentforce-generate

SalesforceAIResearch/agentforce-adlc

Builds, audits, repairs, and deploys Salesforce Agentforce agents written in Agent Script.

117 stars

Machine learning 50

minimal-run-and-audit

Captures smoke test or inference run evidence into standardized repro_outputs files.

481k installs · 486 stars

env-and-assets-bootstrap

Sets up a conservative conda env and records checkpoint, dataset and cache paths for a repro target.

481k installs · 486 stars

ai-research-explore

Runs candidate-only deep learning research exploration with fair comparisons and auditable outputs.

342k installs · 486 stars

analyze-project

Read-only analysis of deep learning repos: model structure, entrypoints and configs, without modifying code.

342k installs · 486 stars

ai-research-reproduction

Reproduces a deep learning repo README-first with minimal, evidence-logged runs and documented deviations.

342k installs · 486 stars

explore-code

Implements exploratory model changes on an isolated branch with rollback-aware records.

341k installs · 486 stars

run-train

K-Dense-AI/scientific-agent-skills

Runs a selected training command for startup or short-run checks, or resume, logging evidence.

341k installs · 486 stars

aeon

Guides time series ML with aeon for classification, regression, forecasting, anomaly detection and search.

55k stars

experiment-queue

wanshuiyin/Auto-claude-code-research-in-sleep

Queues and runs batches of ML experiments on SSH GPU servers with OOM retry and wave scheduling.

18k stars

book-sft-pipeline

muratcankoylan/Agent-Skills-for-Context-Engineering

Builds author-voice SFT datasets from books: segmentation, prompt templates, LoRA training and evaluation.

18k stars

axolotl

Guides LLM fine-tuning with Axolotl YAML configs, covering LoRA, QLoRA, DPO, KTO, ORPO, GRPO and FSDP.

14k stars

distributed-llm-pretraining-torchtitan

Guides large-scale LLM pretraining with TorchTitan across many GPUs using 4D parallelism and checkpointing.

14k stars

grpo-rl-training

Guides GRPO reinforcement learning fine-tuning with TRL, including custom reward functions.

14k stars

huggingface-tokenizers

Guides fast HuggingFace tokenization, custom vocabulary training and text-to-token alignment tracking.

14k stars

implementing-llms-litgpt

Guides loading, training and LoRA or QLoRA fine-tuning of LLMs with LitGPT's single-file architectures.

14k stars

mamba-architecture

Explains and applies Mamba selective state-space models for linear-time sequence modeling and fast inference.

14k stars

miles-rl-training

Guides large-scale RL training of MoE models with miles, covering FP8/INT4 and train-inference alignment.

14k stars

nanogpt

Guides training a minimal GPT-2 style model with nanoGPT on Shakespeare or OpenWebText data.

14k stars

peft-fine-tuning

Guides parameter-efficient fine-tuning with PEFT, LoRA and QLoRA for large models on limited GPU memory.

14k stars

rwkv-architecture

Guides using RWKV, an RNN-style transformer hybrid with linear inference cost, for training and generation.

14k stars

sentencepiece

Guides SentencePiece training and use for language-independent BPE or Unigram tokenization on raw text.

14k stars

sparse-autoencoder-training

Guides training and analyzing sparse autoencoders with SAELens to find interpretable model features.

14k stars

transformer-lens-interpretability

Guides mechanistic interpretability with TransformerLens: activation caching, hooks and activation patching.

14k stars

technology-selection

dotnet/skills

Guides choosing and implementing ML.NET, Microsoft.Extensions.AI, ONNX or local LLM features in .NET 8+.

6k stars

scikit-learn

foryourhealth111-pixel/Vibe-Skills

Guides scikit-learn workflows for classification, regression, clustering, tuning and pipelines.

5k stars

LQF_Machine_Learning_Expert_Guide

Guides ML modeling with baselines, evaluation, overfitting diagnosis, and critical review of your approach.

4k stars

dstack-presets

dstackai/dstack

Creates and manages dstack inference-optimization presets, pointing to external dstack docs for syntax.

2k stars

finetuning

evo-hq/evo

Guides choosing and diagnosing a fine-tuning method such as SFT, DPO, or RL based on reward shape.

1k stars

evaluating-llms-harness

OpenRaiser/NanoResearch

Runs LLMs through 60+ academic benchmarks such as MMLU, GSM8K and HumanEval using a standard harness.

1k stars

huggingface-accelerate

OpenRaiser/NanoResearch

Adds multi-GPU and mixed precision training to PyTorch scripts via Hugging Face Accelerate's unified API.

1k stars

nanoresearch-experiment

OpenRaiser/NanoResearch

Generates a runnable Python experiment skeleton with data, models, training and evaluation from a blueprint.

1k stars

matlab-classify-tabular-data

Guides MATLAB tabular classification: prepare data, train classifiers, and compare accuracy with statistical tests.

1k stars

matlab-deploy-embedded-ai

Deploys neural networks to embedded targets with MATLAB, Simulink and Embedded Coder code generation.

1k stars

matlab-engineer-tabular-features

Selects and engineers features for single-response MATLAB classification or regression models.

1k stars

matlab-import-external-ai-model

Imports PyTorch, ONNX and Keras models into MATLAB as dlnetwork objects and checks numerical match.

1k stars

matlab-interpret-machine-learning-model

Explains tabular MATLAB models using feature importance and per-prediction explanations.

1k stars

matlab-train-network

spencermarx/open-code-review

Trains, evaluates and exports MATLAB neural networks with current dlnetwork APIs and migrates legacy code.

1k stars

optim-agent

Optim-Agent/optim-agent

Tunes configurable parameters against a measured score by reading the project, proposing trials and recording results.

798 stars

agent-lightning

coco-research/coco

Guides training and prompt optimization of AI agents with Microsoft's Agent Lightning using tracing and RL.

795 stars

compare

fcakyon/phd-skills

Compares training runs at the same epoch across trackers, separating proxy metrics from downstream targets.

438 stars

dataset-curation

fcakyon/phd-skills

Analyzes dataset class balance, bias and fairness, and plans stratified samples or data collection.

438 stars

anymove-lora

patch-gg/anymove

Trains a LoRA on fal's MiniMax H3 from captured motions, then A/B tests it against plain H3.

431 stars

anymove-reader

patch-gg/anymove

Renders labelled poses for one creature, fine-tunes a DINOv2 joint reader and checks its held-out error.

431 stars

AgentDB Learning Plugins

Creates and trains agent learning plugins with AgentDB's reinforcement learning algorithms such as Q-Learning.

382 stars

andrej-karpathy

Applies Andrej Karpathy's mental models to neural net training, ML debugging and LLM app design.

287 stars

andrew-ng

Applies Andrew Ng's views on task-level automation, data-centric ML and cheap AI prototyping.

287 stars

christopher-manning

Applies Christopher Manning's view of NLP as a domain science, with modular design and adaptable models.

287 stars

david-silver

lillyguisnet/sam3-segmentation

Applies David Silver's reinforcement learning view: self-play, search, and learning from experience over human data.

287 stars

laya-integration

wdobry/laya-playground

Guides adding the open-source Laya model to classify, route, score or guard text without an LLM call.

193 stars

sam3-segmentation

Runs Meta's SAM 3 locally on an NVIDIA GPU to mask, outline, cut out or count objects in images.

113 stars

Data engineering 23

agent-sources

ccusage/ccusage

Explains where each agent stores logs, how records map to tokens and costs, and which reports each agent supports.

19k stars

kubesphere-fluid

kubesphere/kubesphere

Installs Fluid in KubeSphere and manages Datasets, cache runtimes, DataLoad warming and troubleshooting.

17k stars

nemo-curator

Guides GPU-accelerated LLM training data curation: deduplication, quality filtering and PII redaction.

14k stars

ray-data

Guides distributed ML data processing with Ray Data: streaming preprocessing and batch inference.

14k stars

data-cleaning

genspark-ai/genoffice

Sets cleaning rules, deduplicates and standardises messy tables, and logs every change in an Excel workbook.

11k stars

format-specific-extraction

Describes how Xberg's Office, PDF and archive extractors parse files within security budgets.

9k stars

ocr-pipeline-and-quality

Covers Xberg OCR backends, preprocessing, caching, page acceptance and cross-backend quality rules.

9k stars

pdf-backends

Explains how Xberg picks the native or Pdfium PDF backend and its rendering and OCR fallback.

9k stars

antinet-doc-parse

anbeime/skill

Parses PDF, Word and Excel files into Markdown with metadata and a confidence score, using fallback parsers.

9k stars

arcgis-to-portaljs

Migrates an ArcGIS Hub open data site to PortalJS as PMTiles and GeoParquet with a parity report.

2k stars

portaljs-add-dcat

Adds DCAT catalog feeds in JSON-LD, Turtle and RDF/XML to a PortalJS portal so open-data portals can harvest it.

2k stars

portaljs-add-geo

Converts a geo file into PMTiles and GeoParquet, uploads them to R2, and registers the dataset in a PortalJS portal.

2k stars

portaljs-add-resource

codejunkie99/agentic-stack

Adds a second file, such as a data dictionary or extra period data, to an existing PortalJS dataset.

2k stars

data-flywheel

Turns human-approved, redacted agent runs into local traces, eval cases, and training-ready JSONL files.

2k stars

dbt_ingest

Maps dbt model and source YAML, including column tests, into ktx semantic-layer overlays and column notes.

2k stars

historic_sql_table_digest

Turns one historic SQL table usage file into a concise usage summary recorded as table usage evidence.

2k stars

ingest_triage

Classifies and resolves conflicts found during bundle ingest, such as duplicates, contradictions and evictions.

2k stars

live_database_ingest

vstorm-co/pydantic-deepagents

Turns a live database schema snapshot into semantic-layer sources, joins and wiki notes from table comments.

2k stars

data-formats

Guides the agent to inspect unknown data files by magic bytes and encoding, then parse them safely.

1k stars

executing-spark

ancoleman/ai-design-components

Runs PySpark or Python on Fabric Spark through temporary Livy sessions without creating a notebook.

1k stars

ai-data-engineering

Builds data pipelines for AI systems: RAG ingestion, embeddings, feature stores and orchestration.

526 stars

architecting-data

ancoleman/ai-design-components

Guides data platform design: lake vs warehouse vs lakehouse, data modeling, medallion layers and governance.

526 stars

csv-to-sif

AlpacaLabsLLC/skills-for-architects

Converts a product-library CSV into a SIF schedule for dealer systems, checking field mappings against the vendor spec.

397 stars

Data analysis 72

analytics

coreyhaines31/marketingskills

Plans and checks analytics tracking, including GA4, tag managers, UTM parameters and event tracking plans.

87k installs · 57k stars

anndata

K-Dense-AI/scientific-agent-skills

Reads, writes and manipulates AnnData .h5ad and Zarr objects for single-cell and scverse workflows.

55k stars

ab-test-analysis

phuryn/pm-skills

Analyzes A/B test results for significance, sample size, and confidence intervals, then recommends ship, extend, or stop.

28k stars

cohort-analysis

phuryn/pm-skills

Analyzes user retention and feature adoption by cohort, producing curves, trends, and segment insights from engagement data.

28k stars

ce-product-pulse

EveryInc/compound-engineering-plugin

Builds a read-only, time-windowed product report of usage, performance, and errors from configured sources.

26k stars

google-analytics-admin-api-basics

google/skills

Guides enabling the Analytics Admin API via gcloud and managing accounts, properties, and data streams.

22k stars

google-analytics-data-api-basics

google/skills

Guides enabling the Analytics Data API via gcloud and running custom reports on property data.

22k stars

analyze-results

wanshuiyin/Auto-claude-code-research-in-sleep

Computes statistics and comparison tables from ML experiment results, then writes findings and next experiments.

18k stars

data-analyst

genspark-ai/genoffice

Cleans raw data, builds formula-driven Excel summaries and charts, and states conclusions in plain language.

11k stars

kpi-dashboard

genspark-ai/genoffice

Builds a KPI dashboard workbook with metric definitions, target versus actual, attainment rates and trend charts.

11k stars

hive.chart-creation-foundations

aden-hive/hive

Teaches the agent to render charts with chart_render in a BI-grade style, choosing ECharts or Mermaid.

11k stars

xberg-extracting-keywords

gosom/google-maps-scraper

Extracts keywords with YAKE or RAKE, detects document language and creates embeddings for search.

9k stars

yao-ocr

YaoApp/yao

Extracts text, tables, and structured fields from images and PDFs using VLM or traditional OCR providers.

8k stars

google-maps-scraper

Runs the open-source Google Maps scraper in Docker to collect business leads, contacts and reviews locally.

7k stars

lieflat-charts

larashero3-dotcom/lieflat-charts

Generates HTML charts, slide decks and full-page reports from data, using built-in templates and color presets.

6k stars

sn-da-excel-workflow

Runs an end-to-end Excel analysis workflow: sheet counts, cleaning, filtering, aggregation and exports.

6k stars

sn-da-image-caption

Extracts data from chart and table images by captioning them with a vision model, then structures it.

6k stars

sn-da-large-file-analysis

Handles Excel datasets over 10k rows with streaming reads, Parquet conversion, chunking and memory limits.

6k stars

exploratory-data-analysis

Analyzes scientific data files in 200+ formats and writes a markdown report on structure, quality and statistics.

5k stars

matplotlib

Guides fine-grained Python matplotlib plotting and exports publication-ready PNG, PDF and SVG figures.

5k stars

networkx

Guides building, analyzing, and plotting graphs and networks in Python with NetworkX.

5k stars

seaborn

foryourhealth111-pixel/Vibe-Skills

Guides statistical plots with seaborn for distributions, relationships and categories from pandas data.

5k stars

env-management

aipoch/open-science

Diagnoses missing notebook packages and installs Python or R packages through the managed package tool.

5k stars

claude-dashboard

JimLiu/baoyu-design

Builds a local data dashboard from datasets, with a sources panel showing where each number comes from.

5k stars

alpha-vantage

Fetches stock, forex, crypto, and economic data plus technical indicators from the Alpha Vantage API.

4k stars

data-quality-checker

tradermonty/claude-trading-skills

Flags price scale, ticker, date, allocation and unit errors in market documents before publishing.

3k stars

scipilot-figure-skill

Haojae/scipilot-figure-skill

Profiles scientific data, recommends a chart type, then draws journal-ready figures in matplotlib and plotly.

3k stars

apify-ultimate-scraper

apify/agent-skills

Runs Apify scrapers across social, search, maps, and review platforms for research and monitoring.

2k stars

portaljs-check-data-quality

codejunkie99/agentic-stack

Runs a read-only quality audit on a CSV or TSV file, reporting nulls, types, duplicates and date issues.

2k stars

data-layer

Builds a local terminal dashboard of agent activity, cron runs, token use, and cost across coding tools.

2k stars

app-analytics

appeeky/aso-skills

Guides setup, interpretation and improvement of app analytics, funnels, attribution and KPIs.

2k stars

chart-annotation

ginlix-ai/LangAlpha

Guides the agent to annotate price charts with levels, patterns and events using annotation tools.

2k stars

huashu-data-pro

alchaincyf/huashu-skills

Handles data analysis, Excel formulas, ROI reviews, charts, reports and slides, with an expert perspective.

2k stars

last30days-cn

Jesseovo/last30days-skill-cn

Researches Chinese platforms' last 30 days of posts into cited reports, and shows today's hot-search board.

2k stars

ai-listening

zhuyansen/new-term-radar

Makes the agent collect AI models, products and repos first seen in the last 24 hours across several sources.

2k stars

historic_sql_patterns

Finds recurring cross-table query intents in historic SQL shards and records them as typed pattern evidence.

2k stars

looker_ingest

Extracts metric formulas and field meanings from staged Looker dashboards and explores into ktx memory.

2k stars

lookml_ingest

daymade/claude-code-skills

Maps LookML views, measures and explores into ktx semantic layer sources, with three worked examples.

2k stars

data-visualization-discipline

Applies visualization judgment on conclusion, aggregation, chart type and color before drawing any chart.

2k stars

quick-calculator

alibaba/skill-up

Makes the agent compute arithmetic, unit conversions and formulas with labeled results, steps and checks.

1k stars

create-geo-charts

onvoyage-ai/gtm-engineer-skills

Creates charts and tables with text summaries and JSON-LD so AI engines can parse and cite the underlying data.

1k stars

data-forms

huytieu/COG-second-brain

Chooses a chart or diagram form for a dataset and applies encoding rules so the finding reads at a glance.

1k stars

matlab-analyze-reliability

Guides MATLAB reliability analysis: life distribution fits, MTTF, B10, survival curves and bootstrap intervals.

1k stars

matlab-fit-curve

rorkai/app-store-connect-cli-skills

Guides interactive curve and surface fitting in the MATLAB Curve Fitter app and exports results.

1k stars

scenario-admin-analytics

scenario-labs/skills

Answers Scenario usage questions and builds CSV, Markdown, PDF or dashboard reports from MCP data.

1k stars

asc-analytics-reports

Finds, downloads and checksum-verifies App Store Connect analytics report segments before analysis.

1k stars

creating-good-dashboards

Guides the agent to design readable dashboards and reply to comments without resolving them.

1k stars

deneb-visuals

Guides writing Vega and Vega-Lite specs as Deneb visuals in Power BI PBIR reports.

1k stars

python-visuals

Guides writing matplotlib or seaborn scripts for Python visuals in Power BI PBIR reports.

1k stars

querying-fabric

Queries Fabric and Power BI semantic models with DAX to feed dashboards, reusing existing model measures.

1k stars

r-visuals

Creates R ggplot2 script visuals for Power BI PBIR reports, with the required print step and data rules.

1k stars

svg-visuals

Builds inline SVG charts and indicators in Power BI via DAX measures with ImageUrl data category.

1k stars

using-duckdb