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. Each candidate is checked by a language model from its name, description and topics, and only actual MCP servers are listed (not clients, SDKs, lists or skills). Package descriptions, licences and links are read from each package's own registry (npm, PyPI, crates.io, NuGet, RubyGems; Go and Maven packages are only checked to exist). Packages are found through ecosyste.ms; its metadata (CC BY-SA 4.0) is kept only where the registry could not be read, and packages the registry no longer has are removed. 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.
-
research-osupdated this year
MCP server for reproducible, grounded, citation-verified research — sub-task pipelines, provenance sidecars, publication-grade visualisation, grounded reasoning, and self-tested dashboards.
pypi MIT -
19 research Agent Skills, a reproducible NHANES study, and RigorMed MCP.
NOASSERTION ★ 1 -
FigOps: reproducible research figure pipelines and MCP tooling
Apache-2.0 -
Research-OSgithub.com/VibhavSetlur/Research-OSupdated this year
An AI-native operating system for reproducible research. Features citation-verified analysis, audited provenance, and automated data integrity pipelines.
MIT ★ 1 -
Evidence-first, self-hosted quantitative research workbench with reproducible backtests, factor diagnostics, paper simulation, and grounded AI.
MIT ★ 1 -
Experiment memory for research agents: register attempts, capture outcomes, and report reproducibility through MCP.
-
Give Claude Code real hands on MATLAB & Simulink (MathWorks Agentic Toolkit MCP), then take your simulations to a reproducible paper. Install for macOS/Windows/Linux, ~80 Simulink traps, skills, research template.
NOASSERTION -
Python MCP backend for source-grounded research across patents, publications and web sources, with deterministic scoring, SQLite provenance and reproducible offline evaluation tooling.
MIT -
Guarded MCP and CLI toolkit for reproducible public-source accountability research
MIT ★ 1 -
Verified CPU-only Vendi Score MCP tools: unchanged research functions, real stdio calls and reproducible scientific evidence.
MIT -
Self-hosted portfolio journaling and agent-driven stock research: reproducible experiments, backtesting, durable jobs, API/CLI/MCP extensions, and multi-currency tracking.
MIT ★ 47 -
Open-source research assistant and MCP audit toolkit for systematic reviews: cited research reports, PRISMA tracking, evidence and claim checks, and reproducible exports with human oversight.
npm MIT ★ 112 -
A local MCP server for reproducible quantitative research, backtesting, and Binance Spot Testnet paper trading.
MIT -
Experimental Rust MCP search server with source extraction, evidence-bound research and reproducible evaluations
-
Hybrid retrieval and RAG over research files, with source tracing, reviewable file operations, MCP tools, and reproducible ranking benchmarks.
Apache-2.0 -
Agent-driven ML researcher in a secure, reproducible KVM lab: five standalone scientific-state MCP servers within per-user VMs with driver / trusted / hostile zone isolation
Apache-2.0 ★ 8 -
A live research world where autonomous AI agents challenge, reproduce and audit each other's work. Agents run on their owners' machines; keys never leave. Open protocol, auditable knowledge genealogy, TEST-only economy.
MIT ★ 2 -
Mobile-first genomic intelligence platform: 607,051 SNP calls ingested into EU-hosted Cloudflare D1, accessed through OAuth-secured MCP tools. Integrates GWAS Catalog and Ensembl research, reference-allele validation, reproducible tests, and explicit boundaries for clinical inference.
-
QuantLab Pro is a quantitative ML research platform featuring six specialized agents, MCP-based tool integration, leakage-resistant time-series modeling, PyTorch experiments, backtesting, and reproducible MLOps tracking using Python, Streamlit, and SQLite.
MIT -
Human-guided research workflows: visual questions, reusable code, evidence comparisons and reports with MCP.
MIT ★ 1 -
中文自然语言驱动 COMSOL:Java 官方进度窗口 / MCP 原生 GUI,交付完整可视 MPH
MIT ★ 1 -
Agent-native tool (MCP) for collecting source-traceable terms and evidence spans: recipes, resumable jobs, grounding audits, reproducible datasets.
MIT -
Local, auditable market research for AI agents with exact evidence provenance, claim verification, counterevidence, MCP, and deterministic evidence packs.
MIT ★ 1 -
academic-data-huntergithub.com/Wack520/academic-data-hunterupdated this year
面向研究、竞赛与论文场景的可追溯数据采集与交付工具
MIT ★ 3 -
MCP server for REP v1 research records
npm Apache-2.0 -
Portable pedagogical context for the AI assistant you already use. Local MCP server, persistent profiles and reproducible evaluation.
MIT -
battergithub.com/usepancake/batterupdated this year
The backtesting engine for prediction markets. Every run ships a verifiable, reproducible receipt.
pypi Apache-2.0 -
Deliverable evidence-chain & conformance checks for DSH — turn requirements into an executable spec, trace every number to a provenance ledger, audit PDF/OOXML metadata and archive manifests. MCP-ready.
MIT ★ 1 -
Local Web GUI for reproducible chick/human-eye illumination experiments with guided OpticStudio verification; optional MCP automation.
★ 1 -
Auditable public DHS aggregate research through Python, MCP, Codex and Claude Code — Quantium Insights LLC
NOASSERTION -
Discover GitHub implementations of arXiv research papers. CLI + MCP server for Claude, Cursor, Continue.
pypi MIT ★ 1 -
Local-first research environment: one SQLite ground truth driven by your AI via MCP, with span-anchored verified citations, experiment DAGs, and enforced provenance
MIT -
Research task management, agent conversations, and progress tracking
pypi AGPL-3.0 -
litmusgithub.com/AarushAgarwal-dev/litmusupdated this year
An AI reproducibility auditor that predicts whether a scientific paper's claims will replicate. Deterministic forensic checks, multi-source retrieval, calibrated verdicts, and an MCP server.
MIT ★ 1 -
Open-source Japanese equity quant research stack for J-Quants/local data, factor & ML research, backtesting, portfolio/risk, paper simulation and MCP agents.
AGPL-3.0 ★ 18 -
Unified social-media management — posting, analytics, insights
AGPL-3.0 -
Run unit tests on your manuscript. Check every manuscript claim against its cited evidence. Human-first, AI-second, Zotero-audited, local-first.
pypi Apache-2.0 ★ 1 -
Experimental downstream execution boundary for deterministic authorization and controlled actuation from untrusted or nondeterministic producers.
Apache-2.0 -
mcp4rs-media-gallerygithub.com/MCP4RemoteSensing/mcp4rs-media-galleryupdated this year
Reproducible media-gallery pipeline for MCP4RS: query open remote-sensing sources, record provenance, and render auditable PNG/GIF examples for GitHub, Colab, and Hugging Face.
MIT -
Zenodo-mcpgithub.com/Agostynah/Zenodo-mcpupdated this year
🔬 Model Context Protocol (MCP) server for discoverability, retrieval, citation, and reasoning of open science datasets and publications from Zenodo. AI-native tools for agentic research.
MIT -
Statistical test selection, assumption checks, and result interpretation
AGPL-3.0 -
Genome-editing observability: evidence Supergraph, reusable 3D research lab, exact replay, CLI/MCP, Streamlit and Docker. English / Castellano.
GPL-3.0 -
Trace changed experiment results to the numbers, claims and figures in your LaTeX paper. Offline bilingual review, safe numeric patches and read-only MCP tools.
MIT ★ 200 -
TRACE protocol for auditing decision-provenance in AI-assisted scientific workflows
Apache-2.0 ★ 26 -
Build any paper, prove every claim. IRP turns a research paper into working code and a verifiable, claim-by-claim reproduction report, driven by your AI coding agent over MCP or the CLI.
pypi MIT ★ 3 -
Point your agent at a paper. It reruns the claim and gives you a verdict with receipts, or tells you exactly why it couldn't.
MIT -
A ledger of experimental data and process, kept encrypted and backed by proof. GitHub for code, Hugging Face for models, Provena for verifiable science.
NOASSERTION -
genepattern-mcpgithub.com/genepattern/genepattern-mcpupdated this year
MCP server that lets AI assistants (Claude, Cursor, Copilot) run GenePattern bioinformatics modules, manage cloud jobs, and access genomic data — all through natural language.
BSD-3-Clause ★ 2 -
Stack de bioinformatica e biologia sintetica computacional: servidores MCP proprios para os NIMs da NVIDIA, seis agentes, hooks de proveniencia e 103 testes sem chamada de rede.
-
A database's schema grounded in its data and described in sentences, ready for a text to SQL agent to read
Apache-2.0 ★ 1 -
Reference implementation for the Part II architecture paper (Zenodo 10.5281/zenodo.20374393) — dual-index FAISS memory, 42-opcode quaternionic VM, 5-phase retrieval, 36 tools, MCP server for Claude clients.
MIT ★ 2 -
MCP server that runs RDKit and fpsearch-rs similarity search for an LLM agent. Every tool result is returned with the record of the run that produced it (input and output digests, exit status, completeness), and failed or partial runs are withheld rather than returned.
MIT -
Due diligence on scientific claims, graded against the published record. Claude Code plugin for the Zetesis MCP server, covering omics, clinical trials and AI models. Evidence from Europe PMC, ClinicalTrials.gov, openFDA, NIH RePORTER and SEC EDGAR, with a hard identifier on every source.
MIT -
Evidence-chain whiteboard + MCP server for AI research agents: every claim traces back to the run and code that produced it. 讓 AI agent 做研究時,每個結論都能追到產生它的程式碼。Claude Code · Codex · Cursor · Gemini CLI
npm AGPL-3.0 ★ 501 -
Evidence-first reverse engineering workbench with bounded binary analysis, extensible static adapters, MCP and reproducible local investigations.
MIT -
Auditable research provenance as local-first MCP tools. The claim checker, paper feed, knowledge graph and symbolic math assume nothing about your field — any results JSON or CSV works, from any instrument. The experiment ledger adds W&B/MLflow/Sacred/DVC/Hydra for computational work. Nothing leaves the machine.
MIT ★ 1