data-analysis

K-Dense-AI/scientific-agent-skills

A rebranded, open-standard skills library turning agents into science assistants — Formerly Claude Scientific Skills, this repo now works with any Agent Skills-compatible tool (Cursor, Codex, Pi, Antigravity), not just Claude. It bundles 177 validated skills and 100+ scientific databases for biology, chemistry, medicine, and drug discovery, and powers the new open-source desktop co-scientist K-Dense BYOK.

pandas-dev/pandas

pandas: the Python data toolkit behind countless analyses — pandas is the long-running Python library for labeled, tabular data, offering DataFrame structures, alignment, and statistical tools modeled after R's data.frame. With active CI, NumFOCUS backing, and steady releases on PyPI/conda, it remains a foundational choice for data scientists and analysts working in Python.

metabase/metabase

The easy-to-use open source Business Intelligence and Embedded Analytics tool that lets everyone work with data :bar_chart:

gchq/CyberChef

GCHQ's browser-based toolkit for decoding, hashing, and parsing data — Built by a GCHQ analyst during innovation time, CyberChef runs entirely in the browser and chains operations like XOR, AES, hashing, and X.509 parsing via drag-and-drop recipes. It now ships a Docker image and pre-built container for quick local setup, useful for analysts who need complex data manipulation without scripting.

airbytehq/airbyte

Airbyte widens its data-movement scope to feed AI agents, not just warehouses — Airbyte is an open-source ELT platform syncing APIs, databases and files into warehouses, lakes and now AI applications. It supports self-hosted and Cloud deployments, with a large connector catalog and tooling for data engineers to customize or build new integrations. The MIT/ELv2 licensing keeps the core open while covering long-tail sources.

hicccc77/WeFlow

A fully local WeChat chat viewer, analyzer, and exporter — WeFlow runs entirely on your device to view WeChat chats in real time, decrypt media including Moments photos and Live Photos, block message recall, and generate annual reports with visual stats. It ships as desktop apps for Windows, macOS, and Linux, plus an HTTP API for developers, but requires a custom native decryption component you build yourself.

OpenSenseNova/SenseNova-Skills

Pluggable office skills that extend SenseNova models into full agents — A collection of modular skills—PPT generation, Excel analysis, deep research, HTML experiences and more—each defined via a SKILL.md file following the Agent Skills convention. Built to plug into agent runtimes like OpenClaw and hermes-agent, it's for developers assembling SenseNova-powered office assistants from composable, standalone capabilities.

getnao/nao

Self-hosted agent that turns plain-English questions into SQL insights — nao is a self-hosted framework for building and deploying analytics agents. Data teams define context (data, metadata, docs, tools, MCPs) via the nao-core CLI, then expose a chat UI so business users can query warehouses like BigQuery, Snowflake or Databricks in plain English, with versioning and performance tracking.

graphene-data/graphene

An everything-as-code analytics framework built for AI agents — Graphene gives coding agents a semantic layer and a dashboard file type so they can run SQL analysis and build visualizations with more accuracy and polish than raw Python/JS. Controlled entirely via CLI, it targets teams who want agents to replace or augment traditional BI workflows, with version control and CI baked in.

kobaltgit/StatCraft

A Safe-Rust SPSS Alternative That Also Speaks MCP — StatCraft is an open-source statistical package written in 100% Safe Rust, aimed at researchers, educators, and analysts who need an SPSS-like GUI without crashes or licensing costs. Notably, it ships an MCP bridge so AI agents can drive the GUI directly, making it a candidate tool for both human statisticians and automated analysis workflows.

ashtonckj/Purrrfessorrr-Slothhh

A Pixel-Art Office of Cats Automates Your EDA Boilerplate — This hackathon project turns routine exploratory data analysis into a themed workflow: upload a CSV/TSV/XLSX and five 'specialist cats' handle quality checks, charts, plain-English summaries, and reusable Python/R code. Statistics are computed deterministically; an optional Gemma model only explains results from a summary, never raw data or the math itself.