MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
Composable rule engine for LLM data quality validation with IQR/Z-score anomaly detection & auto-fix pipeline. CLI + MCP ready.
MIT -
An end-to-end AI data pipeline engineered with TimescaleDB, pgvector, and Model Context Protocol (MCP) in a PostgreSQL environment. The system ingests simulated IoT telemetry into hypertables alongside DiskANN vector search to enable autonomous, memory-efficient RAG analysis for AI agents.
-
dagsmithgithub.com/tmcwilliam707/dagsmithupdated this year
Agentic scaffolding for data engineering — turns natural-language source specs into tested Snowflake pipelines (dbt + Airflow + GE) via an LLM tool-use loop over an MCP metadata server.
MIT ★ 1 -
PocketMockstergithub.com/KeigoShimadaCC/PocketMocksterupdated this year
Monster-catching RPG built from scratch in TypeScript + HTML5 canvas, no game engine. Data-driven, scalable content pipeline (creatures, moves, maps, tiles) and an AI-agent debugger: Codex/Cursor agents play the game live over MCP to hunt bugs and file findings reports.
-
-mon-projet-Data-Engeneering-Polargithub.com/Didier2024-hue/-mon-projet-Data-Engeneering-Polarupdated this year
Built a complete Data Engineering pipeline connecting the Polar AccessLink API to InfluxDB for time-series storage, Grafana for real-time dashboards, and a Flask web interface powered by the Anthropic Claude API. A custom MCP server exposes live training data to Claude, enabling natural language sports coaching without any ML model retraining.
MIT -
Fitness Analyst AI — Apple Health / HealthKit data pipeline, analysis engine and MCP server. Python owns the truth; the model narrates.
MIT ★ 1 -
cutiblegithub.com/plokdalberb-byte/cutibleupdated this year
Agent-native montage engine: Timeline-as-Data, verb API, deterministic FFmpeg render, QC loop, MCP server, ingest pipeline, multi-agent swarm, VLM perception, Remotion, OTIO bridge, render farm, REST API, SDK.
MIT ★ 2 -
End-to-end ML toolkit automating the complete workflow — data profiling, preprocessing, feature engineering, model selection, training, validation, explainability, drift monitoring, fairness checks, and interactive HTML reporting. Includes a full CLI for zero-code ML pipelines. Python 3.10+.
pypi MIT ★ 16 -
omega-tools-mcpgithub.com/ranapratapmajee/omega-tools-mcpupdated this year
A production-grade Model Context Protocol (MCP) server engineered for LLM agents. Features highly modular, XML-isolated toolkit pipelines that eliminate formatting bias across custom data lookups and external API actions. Ready for standard local Stdio IDE workflows or scalable, multi-container SSE Docker network meshes.
-
Distributed Data Processing Pipeline for MCP.
AGPL-3.0 ★ 33 -
Creator growth intelligence lakehouse: quota-aware YouTube ingestion, DuckLake bronze, dbt marts, Dagster orchestration, a five-page Streamlit product, 10 SLO checks, and a governed MCP metric server.
MIT -
A sovereign, local-first MCP-Native ETL System with Hot/Cold tiering, PII scrubbing, ID stitching, Parquet buffers, and data warehouse exports.
pypi MIT -
Interactive data systems lab and MCP-native runtime copilot for explainability, regression testing, and operational diagnostics.
oci MIT -
Automated dataset reverse engineering framework — 6-stage analysis pipeline, LLM-enhanced cost modeling & 23+ production documents. CLI + MCP ready.
MIT -
Census ACS to DuckDB to tested dbt marts to governed metrics, served by a read-only MCP server AI agents can query safely.
NOASSERTION -
aide-mcp-servergithub.com/NetApp/aide-mcp-serverupdated this year
MCP server for NetApp AI Data Engine
pypi BSD-3-Clause -
Your entire data pipeline. One platform
pypi AGPL-3.0 -
Root-cause diagnosis for AWS Glue, MWAA (Airflow), Athena and Lake Formation failures. CLI and MCP server, with an optional Claude-on-Bedrock reasoning step.
MIT -
Declarative ETL pipelines in YAML, validated before they run. CLI, REST API, visual Studio, MCP server.
AGPL-3.0 ★ 7 -
Diagnostic multi-agent system for Databricks PySpark job failures: a log-parser -> classifier -> patch-proposer chain built with Claude Code subagents, hooks, and a custom MCP server, gated behind human approval with a full evidence trail.
MIT -
An open-source, AI-powered data engineering workspace for automated SQL/SAS lineage discovery, dependency visualization, column usage analysis, impact assessment, and developer productivity.
Apache-2.0 ★ 2 -
PR-time dbt lineage, impact analysis and rebuild plans — no warehouse, no Python. CLI, TUI, GitHub Action and MCP server, written in Rust.
pypi MIT ★ 23 -
dingdawg-data-agentgithub.com/dingdawg/dingdawg-data-agentupdated this year
Governance receipts for AI agents — data analysis with query optimization, anomaly detection, pipeline monitoring, and built-in PII/GDPR/CCPA detection. Every operation signed and receipted for EU AI Act, Colorado AI Act (SB 24-205), and NIST AI RMF audit
npm BUSL-1.1 -
Open-source ETL/ELT you deploy on your own servers or cloud. Built on DuckDB: no-code/low-code visual pipelines or SQL, 426 components, dbt, CDC, data quality, reverse ETL, lineage, MCP for AI agents. No vendor cloud, no per-row billing.
pypi Apache-2.0 ★ 1,575 -
Orchestration for AI data pipelines, without the platform. A fast Rust core drives flows, schedules, backfills, retries, and a live UI. Agentic-ready: MCP tools, agent-attributed runs, human-in-the-loop approvals. One wheel, no runtime dependencies.
NOASSERTION ★ 1 -
Production-ready MCP server for Apache DolphinScheduler (小海豚) — lets AI agents build, schedule, run and recover workflows
go Apache-2.0 ★ 24 -
DBConvert Streams: Database IDE, Federated SQL, Real-time CDC & AI assistants via MCP — explore, query, and replicate data across databases and files
npm MIT ★ 24 -
coalesce-mcpgithub.com/jarredr092699/coalesce-mcpupdated this year
MCP server for Coalesce data pipeline investigation
pypi MIT -
work-feed-mcpgithub.com/jaeyeopme/work-feed-mcpupdated this year
Dockerized Python data-ingestion engine with SQLite storage and MCP tool access.
MIT -
The Data Control Plane for AI Agents — acquire, validate, and land data over MCP with scoped keys, approval gates, and an audit log. Self-hosted, open source.
pypi Apache-2.0 ★ 14 -
PostgreSQL port of a Snowflake AI analytics lab: dbt, row-level security, and a DeepSeek agent + MCP server with pgvector/full-text search.
Apache-2.0 -
Seed-to-scale LLM synthetic data engine with auto-detected templates, schema validation & quality-diversity optimization. CLI + MCP ready.
MIT ★ 1 -
FRED → PostgreSQL ETL pipeline, deployed on AWS Lambda/RDS in a custom VPC, with an MCP server for AI assistants
-
Open-source AI agent that patrols your data, fixes only what it can prove against the source, and sends the rest to a person with the evidence. Works with Snowflake, dbt and DuckDB, runs as a GitHub Action, and speaks MCP.
pypi MIT ★ 5 -
Chicago Divvy bike-share availability on Google Cloud: a Cloud Run pipeline collects live data every 15 minutes into BigQuery, dbt models and checks it hourly, and a hosted AI agent (Strands + MCP) answers plain-English questions from read-only data tools. Complete code private, only shareable part published in this repo
NOASSERTION ★ 2 -
Business data pipelines, modeling and MCP access through Claude and ChatGPT. Active with two brands and used for internal projects.
-
Agentic ETL pipeline for FOIA & public comments — LangGraph, MCP, PII vault, Medallion warehouse (DuckDB/BigQuery)
Apache-2.0 -
Trail running analytics: Garmin FIT to DuckDB medallion pipeline, exposed to Claude through an MCP server
MIT -
Multi-source async competitive intelligence engine for AI training data ecosystems with watermark-driven incremental scanning & anomaly detection. CLI + MCP ready.
MIT ★ 3 -
Lightweight self-hosted workflow scheduler in a single Go binary — an open-source Airflow/Azkaban alternative. Embedded SQLite, DAGs (cron + dependencies + backfill), polyglot tasks (shell/Python/SQL/JAR/HTTP), web console, REST API, and a built-in MCP server for AI agents. Linux & macOS.
MIT ★ 2 -
Enterprise device analytics with a governed AI assistant: DuckDB medallion pipeline, read-only MCP tools for Claude, evals, human-approved briefings and a live React app. Synthetic data.
-
The reproducible data-cleaning engine for Python. Profile messy CSVs, detect issues, generate a YAML recipe (with or without an LLM), and replay it with zero AI calls.
Apache-2.0 ★ 5 -
Stock-Market Research Assistant - Databricks AI Capstone | Pipeline Spark, RAG com pgvector, MCP Agent, Lakebase
NOASSERTION -
Find, explain, and safely fix bad data. Local-first data quality for files, databases, CI pipelines, and AI agents.
pypi Apache-2.0 ★ 2 -
The safety check for data changes in the AI era: data diffs, Unity Catalog blast radius, policy verdicts and rollback plans for Databricks pipelines and AI agents.
pypi Apache-2.0 ★ 3 -
Readiness system that lets AI agents build your BI pipeline but never approve it -- evidence-gated source profiling, medallion mappings, metric contracts, and Power BI delivery. Offline demo in 60 seconds: uvx --from seshat-bi seshat demo init
npm Apache-2.0 ★ 2 -
Pipeline ETL de datos públicos de la CNMC sobre logística de hidrocarburos: historial mensual, CSV/Parquet, API REST y servidor MCP.
MIT -
STI surveillance in the US: weekly CDC NNDSS + Census pipeline (Neon, dbt), composite Power BI model with RLS/OLS, free static demo and read-only remote MCP
MIT -
@einlogic/mcp-fabric-apiactive
MCP server for Microsoft Fabric REST APIs — enables data engineers and data analysts to use AI assistants beyond Copilot to build and manage Fabric components
npm AGPL-3.0 -
Case study: What's Up In, free event calendars for 57 cities (~440k upcoming events) with a public MCP server. Production engine is private; open MCP client and dedupe code.
MIT -
End-to-end data pipeline: S3 → Snowflake → dbt → Airflow, gated by GitHub Actions CI/CD. Extended with an AI agent layer: MCP server, Cortex vector search, agent-diagnosed bug fix shipped through CI.
MIT -
Multi-source AP invoice pipeline on Snowflake: Dynamic Tables, a Cortex Agent on a semantic view with tested stored-procedure tools and answer-amount checks, MCP for Claude, a BR-004 data-quality guardrail and a CI-gated native agent evaluation, with preserved live evidence