Skip to main content

MatRouter

MatRouter logo

English | 简体中文

MatRouter is an agent-native routing and aggregation substrate for materials data. It turns an explicit data need into executable routes across heterogeneous sources, then assembles returned records, structures, properties, artifacts, provenance, and execution outcomes into one EvidenceBundle.

The Agent (LLM + MCP + Skill) chooses source and record scope and owns planning, retries, scientific interpretation, and prose. MatRouter owns capability truth, source-qualified execution, normalization, provenance, and bundle integrity. It preserves the context needed for downstream reasoning without turning retrieval into a scientific verdict.

Install

MatRouter requires Python 3.11 or newer.

# Python dependency
uv add --prerelease=allow matrouter

# Standalone MCP command
uv tool install --prerelease=allow matrouter

From a checkout:

git clone https://github.com/MatRouter/matrouter.git
cd matrouter
uv sync
uv run matrouter

Python

from matrouter import create_router

with create_router() as router:
    print(router.get_evidence_capability_document())

EvidenceBundle is the core output and the sole record of what was requested, attempted, returned, or missing. The single MCP surface also exposes explicit methods for exact-input structure matching, frame-safe phase-diagram calculation/plotting, and rendering acquired band-structure or DOS artifacts. Retrieval never renders implicitly. Method results retain their exact inputs, parameters, provenance, and limitations; they are not general scientific verdicts.

Agent plugin

Codex and Claude Code can install the bundled plugin; other MCP hosts can run matrouter as a local stdio server and use the portable Skill when supported. See the user guide for the current installation and configuration procedure.

The sole distributable Skill source is plugins/matrouter/skills/matrouter/.

Scope

MatRouter does not mirror upstream databases, infer missing values, impose a fixed research workflow, embed scientific assessment rules, or decide scientific truth. OPTIMADE remains complementary: OPTIMADE standardizes provider-side common records; MatRouter orchestrates native and OPTIMADE sources on the Agent side and preserves their source-specific context in a common bundle.

Documentation

Contact

License

MatRouter is licensed under Apache-2.0; see LICENSE and NOTICE. Retrieved data remains subject to each upstream provider's terms, licenses, attribution requirements, and access restrictions.

Metadata

Release files for matrouter 0.11.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for matrouter 0.11.0
File Size Uploaded Provenance
matrouter-0.11.0.tar.gz 711.4 kB With provenance Details

Built distribution (wheel)

Table of built distributions (wheels) for matrouter 0.11.0
File Interpreter ABI Platform
matrouter-0.11.0-py3-none-any.whl Python 3 none any Details

Total release size: 1.0 MB

Release files / matrouter-0.11.0.tar.gz

Download URL matrouter-0.11.0.tar.gz
Size 711.4 kB
Tags Source
SHA-256 checksum
How to use checksums
537be5b7bf3ab10eac033f1c89ad94e51600d2a9cab9924fa41c06c27b00fbcc
BLAKE2b-256 checksum
How to use checksums
0e4ab17f6d22136c7195365d90bc953823f77c7ba26dfa0c4f7de90b2ebbcc92
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 21, 2026.

Transparency log

Release files / matrouter-0.11.0-py3-none-any.whl

Download URL matrouter-0.11.0-py3-none-any.whl
Size 306.3 kB
Tags Python 3
SHA-256 checksum
How to use checksums
4c63b6addf3d5dfaa2c33659e5dffbd30679ba70302c8eb57cc281786a6778b9
BLAKE2b-256 checksum
How to use checksums
392076830fde10c86d4e79012128f5750bb524c6690e46db18c4823e1258bb70
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 21, 2026.

Transparency log

Security for matrouter 0.11.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

All 2 files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for matrouter-0.11.0.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 21, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page