Skip to main content
Glama

DOE-MCP

MCP servers over public US Department of Energy and national-laboratory data. Every answer names the systems it came from, when they were read, and what it did not cover.

Not affiliated with the US Department of Energy. DOE-MCP is an independent, community-run project. It is not endorsed by, funded by, or connected to DOE, the NNSA, any program office, any national laboratory, or any Power Marketing Administration. "DOE" names the data, not the origin. No answer this software returns is an official government statement.

DOE and its 17 national laboratories publish a large body of public data, across separate catalogs and APIs. This project exposes selected sources by data domain over a declarative source registry, and puts a provenance envelope on every answer so a caller can check where it came from and what it does not cover.

  • Use it. Install, connect a client, ask a question. The literature vertical needs no credential. See examples/README.md for recorded workflows and live queries.

  • Contribute. The most useful contribution is a source manifest, not code. CONTRIBUTING.md has the rules for people; Local coding-agent rules live in AGENTS.md.

  • Understand it. The workflow guide explains the runtime, coverage and evaluation. docs/reference.md lists every tool and argument, read back from the assembled servers.

Status

The shipping servers read DOE publications, energy and environmental data, materials records, microbiome studies and inputs to PNNL’s Global Change Analysis Model (GCAM). Other domains remain registry entries with named next steps. The table below comes from the registry and tool profiles.

Try doe-mcp demo kenya-soil to follow a country code to a model region and its soil carbon recovery period. The global examples also cover Japan’s geothermal resources and Canadian climate simulations.

Servers shipping

doe-research (11 tools, research:default), doe-energy-data (11 tools, energy:default), doe-earth (8 tools, earth:default), doe-materials (7 tools, materials:default), doe-bio (5 tools, bio:default), doe-water (10 tools, water:default), doe-land (10 tools, land:default), doe-emissions (10 tools, emissions:default), doe-power (9 tools, power:default), doe-economy (9 tools, economy:default)

Distinct tools across shipping profiles

76

Skills with bench tasks

29

Registry

88 source manifests, 29 active, 54 organizations (17 national laboratories), 5 server entries and 1 skill pack in the plug-in catalog

Capabilities

42 of 58 in the vocabulary served by an active source

Adapters

basis_sets, curated, daymet, eia_v2, esgf, essdive, facility_status, federal_register, fueleconomy, gcam_csv, json_document, nmdc, opendatasoft, optimade, osti_family, postgrest, sage, self_registry, text_feed, vips

Verified live

Sage / Waggle edge sensor network, Basis Set Exchange, BPA transmission operations feeds, EAGLE-I historical outage datasets, DOE Software Inventory (code.json), DOE Open Data Catalog (data.json), EIA API v2, ESGF search (CMIP6), ESS-DIVE (Environmental System Science Data Infrastructure), Federal Register API (Department of Energy documents), fueleconomy.gov web service, GCAM: Regional economics, GCAM: Emissions, GCAM: Land and agriculture, GCAM: Power technology, GCAM: Water use, Materials Project OPTIMADE endpoint, NERSC Superfacility API (system status), National Microbiome Data Collaborative Runtime API, ORNL DAAC (Daymet, MODIS/VIIRS subsets), Open Energy Data Hub (ORNL), DOE Data Explorer, DOE CODE, DOE PAGES (Public Access Gateway for Energy and Science), OSTI.GOV Search API, US Large-Scale Solar Photovoltaic Database (USPVDB), US Wind Turbine Database (USWTDB)

Not verified

none

Records reachable with no credential

20,153,893

Registry revision

2026-11-09

For Genesis Mission workflows, see the skill and demo guide. Try doe-mcp skills genesis-check-prediction-bands --show for the method and doe-mcp demo load-interval-check for the recorded result.

Related MCP server: FirstData MCP

Install

Not on PyPI yet. Install from a checkout of this repository, from its root:

pipx install .

For development, with the test dependencies:

uv venv --python 3.12 && uv pip install -e ".[all]" --group dev

Then check the install and register the servers with a client:

doe-mcp doctor
doe-mcp configure claude-code        # or claude-desktop, vscode, cursor

doe-mcp configure claude-code --all registers every server at once. Where a source needs a key, it goes in your own credentials file through doe-mcp configure credentials, never into a client config.

Recorded walkthroughs

Inspect the demos: literature evidence, coverage gaps, grid series, facility coordinates, environmental data and materials. The geothermal dataset-card starter and microbiome sample finder include metadata, source evidence and explicit gaps. Run the recorded workflows locally with no key or model call:

doe-mcp demo evidence --output ./evidence-run
doe-mcp demo dataset-card --output ./dataset-draft
doe-mcp demo microbiome --output ./microbiome-samples
doe-mcp demo client
doe-mcp skills

The workflow guide covers live execution, installed skills and optional model evaluation. Recorded examples do not establish current publisher availability or measured agent accuracy.

Try it from the command line

doe-mcp tools call research.search_literature \
  --args '{"query": "perovskite tandem solar cell", "year_from": 2024, "rows": 3}'

The answer is a complete envelope: the records under data, the OSTI collections that were searched under provenance, one evidence entry per record, and coverage.pagination set to truncated, because far more matches exist than the three returned. No credential is involved.

Questions it takes

Question

Tool

What DOE-funded work exists on perovskite tandem cells?

research.search_literature

Where is the dataset behind this paper, and what is its DOI?

research.search_datasets

Has a national laboratory released Python code for lattice QCD?

research.search_software

Can I read the full text, and where?

research.get_record

What does Oak Ridge publish?

registry.lab_crosswalk

I have an old NREL URL that 404s. What happened?

registry.resolve_org

Which DOE catalogs hold geothermal datasets?

discovery.search_all_catalogs

What are DOE's own designated durable data resources?

discovery.list_pure_resources

Does DOE-MCP cover geothermal, and if not, why not?

registry.search_sources

How much electricity did CAISO demand last night?

energy.grid_status

What does this car actually get?

fuel.find_vehicle

What is BPA's load and wind generation right now?

grid.get_bpa_operations

Which EAGLE-I release holds county outage history for 2021, and where is it?

grid.find_outage_history

Which wind turbines and solar farms are inside this bounding box?

facility.find_wind_turbines, facility.find_solar_facilities

What has DOE proposed or finalized on appliance standards this year?

docs.search_rulemakings

Why is the VIPS patent index unavailable?

registry.describe_source (pnnl-vips)

What was the weather at these coordinates every day last summer?

earth.get_daymet_point

What data exists from the NGEE Arctic campaign, and who produced it?

earth.search_datasets, earth.get_dataset

Which climate models ran the historical experiment for surface temperature?

climate.discover_facets, climate.search_cmip

Where are the air-quality sensors near this city, and what is on them?

sensors.find_nodes

Is Perlmutter up, and when is NERSC's next maintenance?

compute.facility_status (in the research:discovery profile)

Is there soil microbiome data from Washington State, and who collected it?

bio.search_biosamples, bio.get_study

Architecture

DOE-MCP layers agent workflows over focused domain servers, typed tools, and GET-only publisher adapters:

graph TD
    User[AI Client / User] -->|Capability Walk| Skills[Skills: Task Guides]
    User -->|MCP Protocol / stdio| Servers[Domain Servers]
    Skills -.->|Composes Capabilities| Servers

    subgraph "Shipping MCP Servers (Profiles: 8-12 tools default, <=20 ceiling)"
        S1["doe-research<br/>literature, datasets, software, rulemakings"]
        S2["doe-energy-data<br/>EIA-930 grid, BPA operations, outage history, facilities"]
        S3["doe-earth<br/>Daymet weather, ESS-DIVE, ESGF, Sage"]
        S4["doe-materials<br/>Materials Project OPTIMADE, Basis Sets"]
        G["GCAM input servers<br/>water, land, emissions, power, economy"]
        S5["doe-bio<br/>NMDC microbiome studies and biosamples"]
    end

    Servers --> S1
    Servers --> S2
    Servers --> S3
    Servers --> S4
    Servers --> G
    Servers --> S5

    subgraph "Core & Adapters (Read-Only GET)"
        Registry[Declarative source registry]
        Adapters[Publisher adapters: pagination, rate limits, dialects]
    end

    S1 & S2 & S3 & S4 & S5 & G --> Registry
    S1 & S2 & S3 & S4 & S5 & G --> Adapters

    Adapters -->|HTTP GET / Dialect Translation| Upstream[Public DOE, National Lab & Federal APIs]

Tool counts per profile, the number of manifests and the adapter list are in the generated status block above.

The envelope

Every tool returns {data, provenance, evidence, coverage, warnings, ...}. The part that matters most is coverage, which has five independent dimensions that are never collapsed into one status:

{"registry": "covered", "execution": "complete", "pagination": "truncated",
 "source_claim": "partial", "result": "hit"}

result: "empty" with registry: "covered" means the searched systems hold no record. registry: "none" means DOE-MCP has no source for that question and the data may well exist. Conflating those two is the most damaging thing a government-data tool can do to someone relying on it.

DOE's additions to the inherited contract: a citation block (several publishers require citation as a condition of use), a required dataset_version (ATB 2024 and ATB 2026 are different data with the same name), typed access recipes for data too big to inline, and three warning codes for confusions this ecosystem actually produces: derived_layer, catalog_vintage, domain_migrated.

Licensing

Code Apache-2.0. The source registry under sources/ is CC0: the inventory of what exists, who stewards it, and what its terms say is more useful unencumbered. Docs CC-BY-4.0. See NOTICE.

Lineage

The packaging shape (one pip package, many focused stdio servers, the doctor/configure CLI) is borrowed with thanks from PNNL's NEPA-MCP. That project does not endorse this one.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      Not graded
      quality
      B
      maintenance
      Enables agents to search, retrieve, and list open-licensed documents with verifiable provenance, attaching sha256, DOI, and OpenTimestamps proof to every response.
      1,531 npm
      -
    • F
      license
      Not graded
      quality
      A
      maintenance
      Enables AI agents and users to turn natural-language data questions into recommended authoritative primary sources and detailed, verified website/API retrieval paths with traceable evidence chains.
      -
    • A
      license
      Not graded
      quality
      B
      maintenance
      A standalone Model Context Protocol server that lets AI clients search satellite scenes, fetch weather and climate series, query geographic layers, and compute vegetation indices and gradual-pattern analyses over environmental time series, returning structured results with source and limit metadata. It exposes 16 tools over stdio, streamable HTTP or legacy SSE and runs its own data retrieval and calculations.
      1
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to search and ask questions across versioned service documentation, retrieve documents and diffs, and find related docs with cited answers over Streamable HTTP, SSE, or stdio.
      296 npm
      MIT