nmdc-runtime 1.3.1__py3-none-any.whl → 2.12.0__py3-none-any.whl

nmdc_runtime/Dockerfile

@@ -0,0 +1,177 @@
+# Note: Most of the steps for the `base` image were copied verbatim from either `fastapi.Dockerfile`,
+#       `dagster.Dockerfile`, or `test.Dockerfile` (indeed, most of the steps were present in all three files).
+#       Reference: https://docs.docker.com/get-started/docker-concepts/building-images/multi-stage-builds/
+#
+# Base this image upon a variant of the official Python 3.10 image that is, in turn,
+# based upon a minimal (slim) variant of the Debian 11 (bullseye) image.
+# Reference: https://hub.docker.com/_/python
+# ────────────────────────────────────────────────────────────────────────────┐
+FROM python:3.10-slim-bullseye AS base
+# ────────────────────────────────────────────────────────────────────────────┘
+
+# Install and upgrade system-level software in a non-interactive way, then delete temporary files.
+# Note: Setting `DEBIAN_FRONTEND=noninteractive` and passing `-y` to `apt-get` makes things non-interactive.
+RUN export DEBIAN_FRONTEND=noninteractive && \
+  apt-get update && \
+  apt-get -y upgrade && \
+  apt-get install -y --no-install-recommends \
+    tini \
+    procps \
+    net-tools \
+    build-essential \
+    git \
+    make \
+    zip \
+    curl \
+    wget \
+    gnupg && \
+  apt-get -y clean && \
+  rm -rf /var/lib/apt/lists/*
+
+# Enable Python's "fault handler" feature, so, when low-level errors occur (e.g. segfaults), Python prints lots of info.
+# Reference: https://docs.python.org/3/using/cmdline.html#envvar-PYTHONFAULTHANDLER
+ENV PYTHONFAULTHANDLER=1
+
+# Configure Git to consider the `/code` directory to be "safe", so that, when a Git repository
+# created outside of the container gets mounted at that path within the container, the
+# `uv-dynamic-versioning` tool running within the container does not fail with the error:
+# > "Detected Git repository, but failed because of dubious ownership"
+# Reference: https://git-scm.com/docs/git-config#Documentation/git-config.txt-safedirectory
+RUN git config --global --add safe.directory /code
+
+# Install `uv`.
+# Reference: https://docs.astral.sh/uv/guides/integration/docker/#installing-uv
+ADD https://astral.sh/uv/install.sh /uv-installer.sh
+RUN sh /uv-installer.sh && \
+    rm /uv-installer.sh
+ENV PATH="/root/.local/bin/:$PATH"
+
+# Install Python dependencies (production dependencies only).
+#
+# Note: We copy only the files that `uv` needs in order to install dependencies. That way,
+#       we minimize the number of files whose changes would invalidate cached image layers
+#
+# Note: We use the `VIRTUAL_ENV` environment variable to specify the path to the Python virtual
+#       environment that we want the `uv` program inside the container to create and use.
+#
+#       Q: Why don't we use `./.venv` in the repository file tree?
+#       A: If we were to do that, then, whenever a developer would mount (via our Docker Compose file)
+#          the repository file tree from their host machine (which may include a `.venv/` directory
+#          created by their host machine) into the container, it would overwrite the Python virtual
+#          environment that the `uv` program inside the container is using.
+#
+#       Q: What is special about the `VIRTUAL_ENV` environment variable?
+#       A: When using `uv`'s `--active` option (as we do in later stages of this Dockerfile),
+#          `uv` determines which virtual environment is active by looking at `VIRTUAL_ENV'. This
+#          is the case, even though the documentation of the `venv` module (in Python's standard
+#          library) specifically says: "`VIRTUAL_ENV` cannot be relied upon to determine whether
+#          a virtual environment is being used."
+#
+#       References:
+#       - https://docs.astral.sh/uv/pip/environments/#using-arbitrary-python-environments (RE: `VIRTUAL_ENV`)
+#       - https://docs.astral.sh/uv/reference/environment/#virtual_env (RE: `VIRTUAL_ENV`, from uv's perspective)
+#       - https://docs.python.org/3/library/venv.html#how-venvs-work (RE: `VIRTUAL_ENV`, from venv's perspective)
+#       - https://docs.astral.sh/uv/concepts/projects/sync/#partial-installations (RE: `--no-install-project`)
+#
+# Note: In the `RUN` command, we use a "cache mount" (a feature of Docker) to cache production dependencies
+#       across builds. This is a performance optimization technique shown in the `uv` docs.
+#       Reference:
+#       - https://docs.astral.sh/uv/guides/integration/docker/#caching (RE: the technique)
+#       - https://docs.docker.com/build/cache/optimize/#use-cache-mounts (RE: the feature)
+#       - https://docs.astral.sh/uv/reference/settings/#link-mode (RE: `UV_LINK_MODE`)
+#       - https://docs.astral.sh/uv/reference/cli/#uv-sync--no-install-project (RE: `--no-install-project`)
+#
+# Note: We use `--compile-bytecode` so that Python compiles `.py` files to `.pyc` files now,
+#       instead of when the container is running. By default, `uv` defers this compilation
+#       to "import time," whereas `pip` (by default) performs it at "install time" (like this).
+#
+# Note: We use `--locked` so that `uv sync` exits with an error if the `uv.lock` file isn't _already_
+#       up to date. By default, `uv sync` would automatically update the lock file if necessary.
+#       Reference: https://docs.astral.sh/uv/reference/cli/#uv-sync--locked
+#
+ENV VIRTUAL_ENV="/venv"
+RUN mkdir -p "${VIRTUAL_ENV}"
+COPY ./pyproject.toml /code/pyproject.toml
+COPY ./uv.lock        /code/uv.lock
+RUN --mount=type=cache,target=/root/.cache/uv \
+    cd /code && \
+    UV_LINK_MODE=copy uv sync --active --no-dev --no-install-project --compile-bytecode --locked
+
+# ────────────────────────────────────────────────────────────────────────────┐
+FROM base AS fastapi
+# ────────────────────────────────────────────────────────────────────────────┘
+
+# Copy repository contents into image.
+COPY . /code
+
+# Install the project in editable mode.
+RUN --mount=type=cache,target=/root/.cache/uv \
+    cd /code && \
+    uv sync --active --no-dev --compile-bytecode --locked
+
+# Use Uvicorn to serve the FastAPI app on port 8000.
+#
+# Note: We include the `--no-sync` option to prevent `uv run` from automatically syncing dependencies.
+#       If it were to sync dependencies at this point, it would install development dependencies, since
+#       we exclude them above, but they are listed in uv's `default-groups` configuration by default.
+#       This is explained at: https://github.com/astral-sh/uv/issues/12558#issuecomment-2764611918
+#
+EXPOSE 8000
+WORKDIR /code
+CMD ["uv", "run", "--active", "--no-sync", "uvicorn", "nmdc_runtime.api.main:app", "--proxy-headers", "--host", "0.0.0.0", "--port", "8000"]
+
+# ────────────────────────────────────────────────────────────────────────────┐
+FROM base AS dagster
+# ────────────────────────────────────────────────────────────────────────────┘
+
+# Copy repository contents into image.
+#
+# Note: This path (i.e. "/opt/dagster/lib/") is hard-coded in a few places in `nmdc_runtime/site/ops.py`. That's why
+#       this image does not store the repository contents in `/code`, unlike the other images in this Dockerfile.
+#
+COPY . /opt/dagster/lib
+
+# Install the project in editable mode.
+RUN --mount=type=cache,target=/root/.cache/uv \
+    cd /opt/dagster/lib && \
+    uv sync --active --no-dev --compile-bytecode --locked
+
+# Move Dagster configuration files to the place Dagster expects.
+ENV DAGSTER_HOME="/opt/dagster/dagster_home/"
+RUN mkdir -p                                             "${DAGSTER_HOME}" && \
+    cp /opt/dagster/lib/nmdc_runtime/site/dagster.yaml   "${DAGSTER_HOME}" && \
+    cp /opt/dagster/lib/nmdc_runtime/site/workspace.yaml "${DAGSTER_HOME}"
+
+# Use Tini to run Dagit.
+#
+# Notes:
+# - The port number (i.e. "3000") is hard-coded in `nmdc_runtime/site/entrypoint-dagit.sh`.
+# - Dagster daemon (versus Dagit) can be launched by overriding the `ENTRYPOINT` defined here.
+#
+# Reference: https://github.com/krallin/tini
+#
+EXPOSE 3000
+WORKDIR /opt/dagster/dagster_home/
+ENTRYPOINT ["tini", "--", "../lib/nmdc_runtime/site/entrypoint-dagit.sh"]
+
+# ────────────────────────────────────────────────────────────────────────────┐
+FROM base AS test
+# ────────────────────────────────────────────────────────────────────────────┘
+
+# Copy all repository contents into image.
+COPY . /code
+
+# Install the project in editable mode, and install development dependencies.
+RUN --mount=type=cache,target=/root/.cache/uv \
+    cd /code && \
+    uv sync --active --compile-bytecode --locked
+
+# Make `wait-for-it.sh` executable.
+RUN chmod +x /code/.docker/wait-for-it.sh
+
+WORKDIR /code
+
+# Ensure started container does not exit, so that a subsequent `docker exec` command can run tests.
+# For an example `docker exec` command, see `Makefile`'s `run-test` target.
+# Such a command should use `wait-for-it.sh` to run `pytest` no earlier than when the FastAPI server is accessible.
+ENTRYPOINT ["tail", "-f", "/dev/null"]

nmdc_runtime/api/analytics.py

@@ -0,0 +1,90 @@
+"""
+Based on <https://github.com/tom-draper/api-analytics/tree/main/analytics/python/fastapi>
+under MIT License <https://github.com/tom-draper/api-analytics/blob/main/analytics/python/fastapi/LICENSE>
+"""
+
+from datetime import datetime
+import threading
+from time import time
+from typing import Dict, List
+
+from starlette.middleware.base import BaseHTTPMiddleware, RequestResponseEndpoint
+from starlette.requests import Request
+from starlette.responses import Response
+from starlette.types import ASGIApp
+from toolz import merge
+
+from nmdc_runtime.api.db.mongo import get_mongo_db
+
+# This is a queue of the "request descriptors" that we will eventually insert into the database.
+_requests = []
+_last_posted = datetime.now()
+
+
+def _post_requests(collection: str, requests_data: List[Dict], source: str):
+    """Inserts the specified request descriptors into the specified MongoDB collection."""
+    mdb = get_mongo_db()
+    mdb[collection].insert_many([merge(d, {"source": source}) for d in requests_data])
+
+
+def log_request(collection: str, request_data: Dict, source: str = "FastAPI"):
+    """Flushes the queue of request descriptors to the database if enough time has passed since the previous time."""
+    global _requests, _last_posted
+    _requests.append(request_data)
+    now = datetime.now()
+    # flush queue every minute at most
+    if (now - _last_posted).total_seconds() > 60.0:
+        # Note: This use of threading is an attempt to avoid blocking the current thread
+        #       while performing the insertion(s).
+        #
+        # TODO: Is there is a race condition here? If multiple requests arrive at approximately
+        #       the same time, is it possible that each one causes a different thread to be
+        #       started, each with a different (and possibly overlapping) set of requests to
+        #       insert?
+        #
+        # TODO: If the insertion fails, will the requests be lost?
+        #
+        # Note: The author of this function said it may have been a "standard" solution copied
+        #       from some documentation. Indeed, the comment at the top of this module contains
+        #       a link to code on which it was based.
+        #
+        threading.Thread(
+            target=_post_requests, args=(collection, _requests, source)
+        ).start()
+        _requests = []  # empties the queue
+        _last_posted = now
+
+
+class Analytics(BaseHTTPMiddleware):
+    def __init__(self, app: ASGIApp, collection: str = "_runtime.analytics"):
+        super().__init__(app)
+        self.collection = collection
+
+    async def dispatch(
+        self, request: Request, call_next: RequestResponseEndpoint
+    ) -> Response:
+        start = time()
+        response = await call_next(request)
+
+        # Use a fallback IP address value (currently an empty string) if we can't derive one from the request.
+        ip_address: str = "" if request.client is None else request.client.host
+
+        # Build a dictionary that describes the incoming request.
+        #
+        # Note: `request.headers` is an instance of `MultiDict`. References:
+        #       - https://www.starlette.io/requests/#headers
+        #       - https://multidict.aio-libs.org/en/stable/multidict/
+        #
+        request_data = {
+            "hostname": request.url.hostname,
+            "ip_address": ip_address,
+            "path": request.url.path,
+            "user_agent": request.headers.get("user-agent"),
+            "method": request.method,
+            "status": response.status_code,
+            "response_time": int((time() - start) * 1000),
+            "created_at": datetime.now().isoformat(),
+        }
+
+        log_request(self.collection, request_data, "FastAPI")
+        return response

nmdc_runtime/api/boot/capabilities.py

@@ -0,0 +1,9 @@
+from nmdc_runtime.api.models.capability import Capability
+import nmdc_runtime.api.boot.workflows as workflows_boot
+
+# Include 1-to-1 "I can run this workflow" capabilities.
+_raw = [item for item in workflows_boot._raw]
+
+
+def construct():
+    return [Capability(**kwargs) for kwargs in _raw]

nmdc_runtime/api/boot/object_types.py

@@ -0,0 +1,126 @@
+from datetime import datetime, timezone
+
+from toolz import get_in
+
+from nmdc_runtime.api.models.object_type import ObjectType
+from nmdc_runtime.util import nmdc_jsonschema
+
+_raw = [
+    {
+        "id": "read_qc_analysis_activity_set",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        # "description": "JSON documents satisfying schema for readqc analysis activity",
+    },
+    {
+        "id": "metagenome_sequencing_activity_set",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        # "description": "JSON documents satisfying schema for metagenome sequencing activity",
+    },
+    {
+        "id": "mags_activity_set",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        # "description": "JSON documents satisfying schema for mags activity",
+    },
+    {
+        "id": "metagenome_annotation_activity_set",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        # "description": "JSON documents satisfying schema for metagenome annotation activity",
+    },
+    {
+        "id": "metagenome_assembly_set",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        # "description": "JSON documents satisfying schema for metagenome assembly activity",
+    },
+    {
+        "id": "read_based_taxonomy_analysis_activity_set",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        # "description": "JSON documents satisfying schema for read based analysis activity",
+    },
+    {
+        "id": "metadata-in",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "metadata submission",
+        "description": "Input to the portal ETL process",
+    },
+    {
+        "id": "metaproteomics_analysis_activity_set",
+        "created_at": datetime(2021, 8, 23, tzinfo=timezone.utc),
+        "name": "metaP analysis activity",
+        "description": "JSON documents satisfying schema for metaproteomics analysis activity",
+    },
+    {
+        "id": "metagenome_raw_paired_end_reads",
+        "created_at": datetime(2021, 8, 24, tzinfo=timezone.utc),
+        "name": "Metagenome Raw Paired-End Reads Workflow Input",
+        "description": "workflow input",
+    },
+    {
+        "id": "metatranscriptome_raw_paired_end_reads",
+        "created_at": datetime(2021, 9, 7, tzinfo=timezone.utc),
+        "name": "Metatranscriptome Raw Paired-End Reads Workflow Input",
+        "description": "workflow input 2",
+    },
+    {
+        "id": "gcms-metab-input",
+        "created_at": datetime(2021, 9, 7, tzinfo=timezone.utc),
+        "name": "Raw GCMS MetaB Input",
+        "description": "",
+    },
+    {
+        "id": "gcms-metab-calibration",
+        "created_at": datetime(2021, 9, 7, tzinfo=timezone.utc),
+        "name": "Raw GCMS MetaB Calibration",
+        "description": "",
+    },
+    {
+        "id": "nom-input",
+        "created_at": datetime(2021, 9, 7, tzinfo=timezone.utc),
+        "name": "Raw FTMS MetaB Input",
+        "description": "",
+    },
+    {
+        "id": "test",
+        "created_at": datetime(2021, 9, 7, tzinfo=timezone.utc),
+        "name": "A test object type",
+        "description": "For use in unit and integration tests",
+    },
+    {
+        "id": "metadata-changesheet",
+        "created_at": datetime(2021, 9, 30, tzinfo=timezone.utc),
+        "name": "metadata changesheet",
+        "description": "Specification for changes to existing metadata",
+    },
+]
+
+_raw.extend(
+    [
+        {
+            "id": key,
+            "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+            "name": key,
+            # "description": spec["description"],
+        }
+        for key, spec in nmdc_jsonschema["properties"].items()
+        if key.endswith("_set")
+    ]
+)
+_raw.append(
+    {
+        "id": "schema#/definitions/Database",
+        "created_at": datetime(2021, 9, 14, tzinfo=timezone.utc),
+        "name": "Bundle of one or more metadata `*_set`s.",
+        "description": get_in(
+            ["definitions", "Database", "description"], nmdc_jsonschema
+        ),
+    }
+)
+
+
+def construct():
+    return [ObjectType(**kwargs) for kwargs in _raw]

nmdc_runtime/api/boot/triggers.py

@@ -0,0 +1,84 @@
+from datetime import datetime, timezone
+
+from nmdc_runtime.api.models.trigger import Trigger
+
+_raw = [
+    {
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "object_type_id": "metadata-in",
+        "workflow_id": "metadata-in-1.0.0",
+    },
+    {
+        "created_at": datetime(2021, 9, 1, tzinfo=timezone.utc),
+        "object_type_id": "metaproteomics_analysis_activity_set",
+        "workflow_id": "metap-metadata-1.0.0",
+    },
+    {
+        "created_at": datetime(2021, 9, 1, tzinfo=timezone.utc),
+        "object_type_id": "metagenome_raw_paired_end_reads",
+        "workflow_id": "metag-1.0.0",
+    },
+    {
+        "created_at": datetime(2021, 9, 7, tzinfo=timezone.utc),
+        "object_type_id": "metatranscriptome_raw_paired_end_reads",
+        "workflow_id": "metat-1.0.0",
+    },
+    {
+        "created_at": datetime(2021, 9, 9, tzinfo=timezone.utc),
+        "object_type_id": "test",
+        "workflow_id": "test",
+    },
+    {
+        "created_at": datetime(2021, 9, 20, tzinfo=timezone.utc),
+        "object_type_id": "nom-input",
+        "workflow_id": "nom-1.0.0",
+    },
+    {
+        "created_at": datetime(2021, 9, 20, tzinfo=timezone.utc),
+        "object_type_id": "gcms-metab-input",
+        "workflow_id": "gcms-metab-1.0.0",
+    },
+    {
+        "created_at": datetime(2021, 9, 30, tzinfo=timezone.utc),
+        "object_type_id": "metadata-changesheet",
+        "workflow_id": "apply-changesheet-1.0.0",
+    },
+    {
+        "created_at": datetime(2022, 1, 20, tzinfo=timezone.utc),
+        "object_type_id": "metagenome_sequencing_activity_set",
+        "workflow_id": "mgrc-1.0.6",
+    },
+    {
+        "created_at": datetime(2022, 1, 20, tzinfo=timezone.utc),
+        "object_type_id": "metagenome_sequencing_activity_set",
+        "workflow_id": "metag-1.0.0",
+    },
+    {
+        "created_at": datetime(2022, 1, 20, tzinfo=timezone.utc),
+        "object_type_id": "metagenome_annotation_activity_set",
+        "workflow_id": "mags-1.0.4",
+    },
+    {
+        "created_at": datetime(2022, 1, 20, tzinfo=timezone.utc),
+        "object_type_id": "metagenome_assembly_set",
+        "workflow_id": "mgann-1.0.0",
+    },
+    {
+        "created_at": datetime(2022, 1, 20, tzinfo=timezone.utc),
+        "object_type_id": "read_qc_analysis_activity_set",
+        "workflow_id": "mgasm-1.0.3",
+    },
+    {
+        "created_at": datetime(2022, 1, 20, tzinfo=timezone.utc),
+        "object_type_id": "read_qc_analysis_activity_set",
+        "workflow_id": "mgrba-1.0.2",
+    },
+]
+
+
+def construct():
+    models = []
+    for kwargs in _raw:
+        kwargs["id"] = f'{kwargs["object_type_id"]}--{kwargs["workflow_id"]}'
+        models.append(Trigger(**kwargs))
+    return models

nmdc_runtime/api/boot/workflows.py

@@ -0,0 +1,116 @@
+from datetime import datetime, timezone
+
+from nmdc_runtime.api.models.workflow import Workflow
+
+_raw = [
+    {
+        "id": "metag-1.0.0",
+        "created_at": datetime(2021, 8, 24, tzinfo=timezone.utc),
+        "name": "Metagenome Analysis Workflow (v1.0.0)",
+    },
+    {
+        "id": "readqc-1.0.6",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Reads QC Workflow (v1.0.1)",
+    },
+    {
+        "id": "mags-1.0.4",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Read-based Analysis (v1.0.1)",
+    },
+    {
+        "id": "mgrba-1.0.2",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Read-based Analysis (v1.0.1)",
+    },
+    {
+        "id": "mgasm-1.0.3",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Metagenome Assembly (v1.0.1)",
+    },
+    {
+        "id": "mgann-1.0.0",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Metagenome Annotation (v1.0.0)",
+    },
+    {
+        "id": "mgasmbgen-1.0.1",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Metagenome Assembled Genomes (v1.0.2)",
+    },
+    {
+        "id": "metat-0.0.2",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Metatranscriptome (v0.0.2)",
+    },
+    {
+        "id": "metap-1.0.0",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Metaproteomic (v1.0.0)",
+    },
+    {
+        "id": "metab-2.1.0",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "Metabolomics  (v2.1.0)",
+    },
+    {
+        "id": "gold-translation-1.0.0",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "GOLD db dump translation",
+        "description": "Transform metadata obtained from the JGI GOLD database.",
+    },
+    {
+        "id": "metap-metadata-1.0.0",
+        "created_at": datetime(2021, 6, 1, tzinfo=timezone.utc),
+        "name": "metaP metadata ETL",
+        "description": "Ingest and validate metaP metadata",
+    },
+    {
+        "id": "metadata-in-1.0.0",
+        "created_at": datetime(2021, 10, 12, tzinfo=timezone.utc),
+        "name": "general metadata ETL",
+        "description": "Validate and ingest metadata from JSON files",
+    },
+    {
+        "id": "test",
+        "created_at": datetime(2021, 9, 9, tzinfo=timezone.utc),
+        "name": "A test workflow",
+        "description": "For use in unit and integration tests",
+    },
+    {
+        "id": "gcms-metab-1.0.0",
+        "created_at": datetime(2021, 9, 20, tzinfo=timezone.utc),
+        "name": "GCMS-based metabolomics",
+    },
+    {
+        "id": "nom-1.0.0",
+        "created_at": datetime(2021, 9, 20, tzinfo=timezone.utc),
+        "name": "Natural Organic Matter characterization",
+    },
+    {
+        "id": "apply-changesheet-1.0.0",
+        "created_at": datetime(2021, 9, 30, tzinfo=timezone.utc),
+        "name": "apply metadata changesheet",
+        "description": "Validate and apply metadata changes from TSV/CSV files",
+    },
+    {
+        "id": "export-study-biosamples-as-csv-1.0.0",
+        "created_at": datetime(2022, 6, 8, tzinfo=timezone.utc),
+        "name": "export study biosamples metadata as CSV",
+        "description": "Export study biosamples metadata as CSV",
+    },
+    {
+        "id": "gold_study_to_database",
+        "created_at": datetime(2023, 2, 17, tzinfo=timezone.utc),
+        "name": "Get nmdc:Database for GOLD study",
+        "description": "For a given GOLD study ID, produce an nmdc:Database representing that study and related entities",
+    },
+]
+
+
+def construct():
+    models = []
+    for kwargs in _raw:
+        kwargs["capability_ids"] = []
+        models.append(Workflow(**kwargs))
+    return models