mostlyright-data 0.17.6__tar.gz → 0.17.7__tar.gz

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mostlyright-data
-Version: 0.17.6
+Version: 0.17.7
 Summary: Mostly Right CLI for reproducible, reviewed Parquet datasets
 Project-URL: Homepage, https://mostlyright.md/
 Project-URL: Documentation, https://mostlyright.md/docs/guides/cli/
@@ -208,9 +208,12 @@ mr-data --help
 
 ## Build a dataset
 
-Register a recipe document with the sources, table schema, transformations and checks:
+Create the dataset, choose its primary category from the fixed vocabulary, then register a recipe document with the sources, table schema, transformations and checks. Use the created dataset ID in the recipe:
 
 ```bash
+mr-data dataset create --name "Dataset title" --json
+mr-data dataset categories --json
+mr-data dataset set DATASET_ID --category climate-environment
 mr-data recipe recipe.json --json
 mr-data run --recipe RECIPE_ID --digest RECIPE_DIGEST --full
 mr-data status RUN_ID

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/README.md

@@ -45,9 +45,12 @@ mr-data --help
 
 ## Build a dataset
 
-Register a recipe document with the sources, table schema, transformations and checks:
+Create the dataset, choose its primary category from the fixed vocabulary, then register a recipe document with the sources, table schema, transformations and checks. Use the created dataset ID in the recipe:
 
 ```bash
+mr-data dataset create --name "Dataset title" --json
+mr-data dataset categories --json
+mr-data dataset set DATASET_ID --category climate-environment
 mr-data recipe recipe.json --json
 mr-data run --recipe RECIPE_ID --digest RECIPE_DIGEST --full
 mr-data status RUN_ID

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mostlyright-data"
-version = "0.17.6"
+version = "0.17.7"
 description = "Mostly Right CLI for reproducible, reviewed Parquet datasets"
 readme = "README.md"
 requires-python = ">=3.13,<3.14"

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/skills/mr-data-build/SKILL.md

@@ -118,7 +118,7 @@ gap rather than doing anything, and `export-hosted-candidate`, which is a backen
 | `mr-data whoami` | Use the compatibility alias for the remotely validated `auth status` answer. |
 | `mr-data clarify` | Say whether one clarifying question may be put to the user right now, and compose the record of their answer. It reaches nothing, asks nobody and waits for nothing. Exits non-zero when no person can answer or when a recipe is already registered. |
 | `mr-data probe` | Ask one registered source one question and print the answer. **Currently unusable:** it reaches `/v3/sessions`, which Studio refuses, and it takes a question identifier that no v4 receipt returns. Do not plan a source inspection around it. |
-| `mr-data dataset` | Bring the dataset page into existence before there is anything on it, then fill it in while somebody watches. `dataset create --name TEXT` mints it and prints the `dataset_id`; `dataset show ID` reads it back; `dataset set ID --name TEXT --topics "a,b,c" --license ID --description-file F` writes the title, the categories, the SPDX licence and the description under the version it was read at, retrying once if somebody else wrote first, and an empty `--topics` or `--license` takes that value off the page; `dataset note ID --heading H --blocks-file B` writes one cell of the decision record that OUTLIVES every run, and `--list` reads it back; `dataset watch ID` follows the page's own event stream; `dataset activity ID --phase P --message TEXT` says what is happening right now, silently, and is never a chat message and never a cell; `dataset publish ID [--mode public|link|private]` says who can read the dataset — `public` lists it in the public directory and serves it at an address anybody can read, `link` serves it at an unlisted address, `private` takes it back to the workspace — and `dataset publish ID --show` reads that back without changing it; `dataset archive ID --confirm-name TITLE` retires the page and frees its title, deleting nothing. |
+| `mr-data dataset` | Bring the dataset page into existence before there is anything on it, then fill it in while somebody watches. `dataset create --name TEXT` mints it and prints the `dataset_id`; `dataset show ID` reads it back; `dataset set ID --name TEXT --topics "a,b,c" --license ID --description-file F` writes the title, the descriptive tags, the SPDX licence and the description under the version it was read at, retrying once if somebody else wrote first, and an empty `--topics` or `--license` takes that value off the page; `dataset note ID --heading H --blocks-file B` writes one cell of the decision record that OUTLIVES every run, and `--list` reads it back; `dataset watch ID` follows the page's own event stream; `dataset activity ID --phase P --message TEXT` says what is happening right now, silently, and is never a chat message and never a cell; `dataset publish ID [--mode public|link|private]` says who can read the dataset — `public` lists it in the public directory and serves it at an address anybody can read, `link` serves it at an unlisted address, `private` takes it back to the workspace — and `dataset publish ID --show` reads that back without changing it; `dataset archive ID --confirm-name TITLE` retires the page and frees its title, deleting nothing. |
 | `mr-data recipe` | Register one recipe document — dataset, question, table plan, sources, transform, checks and units — in one call, and print the identifiers the server derived. `mr-data recipe show ID` reads one back. |
 | `mr-data cover` | Generate and attach one branded 1200×630 dataset cover. Give it the dataset ID and what the image should depict; Studio fixes the model, single-color style, curated random palette, dimensions and storage. |
 | `mr-data run` | Start one run against a registered recipe, named as `--recipe RECIPE_ID --digest RECIPE_DIGEST` — both required, neither positional, and the digest is the bare hex the registration receipt printed. The command accepts `--sample`, `--full`, `--refresh`, `--backfill` and `--window`, and the worker executes `--sample`, `--full` and `--refresh`. A refresh narrows each source that declares a supported address window and merges its returned relation into retained source state; unwindowed sources are conditionally revalidated or re-acquired whole, then the full declared transform runs. `--backfill` and a `--window` on a sample or full are refused before acquisition. `--max-rows` and `--max-source-bytes` remain supported, and each bounds ONE SOURCE rather than the finished table. A large run is held for a spend confirmation, which `--confirm` settles; a sample-first held full requires Studio's interactive human approval instead. `--cancel RUN` stops one. |
@@ -471,16 +471,16 @@ and is settled once, at stage 4.
 Every grain, key, join and cutoff settled after stage 3 is a further `decision` block on the
 DATASET's record — `mr-data dataset note DATASET_ID` — because the fact outlives every run.
 
-**Write the title, the topics, the licence and the description before the first run.** The page has
+**Write the title, primary category, tags, licence and description before recipe registration.** The page has
 been open since stage 1 under a working title. This is where it settles. One call carries the title,
-the topics and the licence, so the title and the chips land together; the description arrives after
+the primary category, tags and licence, so the title and category land together; the description arrives after
 it, through the draft snapshots and the final commit. (The worked example below is a different
 dataset from the Kathmandu one the earlier stages follow, chosen because its numbers are easy to
 count.)
 
 ```sh
 mr-data dataset set DATASET_ID --name "Hourly weather at Prague Airport (LKPR), 2000-2026" \
-  --topics "weather,prague,czech republic,metar,iowa environmental mesonet" \
+  --category climate-environment --topics "weather,prague,czech republic,metar,iowa environmental mesonet" \
   --license "CC-BY-4.0" --json
 ```
 
@@ -504,7 +504,8 @@ interpret them.
 | --- | --- | --- |
 | `--name` | browser tab title (`{name} \| Mostly Right`), the page heading, the JSON-LD `name` | at most 60 characters, or search results cut it |
 | the opening paragraph of DESCRIPTION.md | the meta description, the social card, the JSON-LD `description`, the `Summary:` line answer engines fetch | search engines cut at 300 characters; answer engines cut at 240 |
-| `--topics` | category chips and the category filter on the /datasets directory, JSON-LD `keywords`, the `Topics:` line answer engines fetch, the `topic` filter in the public API and MCP | 3 to 8 topics, each at most 40 characters |
+| `--category` | the primary category and category filter on Community | exactly one fixed category ID |
+| `--topics` | JSON-LD `keywords`, the `Topics:` line answer engines fetch, the `topic` filter in the public API and MCP | 3 to 8 topics, each at most 40 characters |
 
 **Title rules**
 1. Shape: subject, place, period. `Hourly weather at Prague Airport (LKPR), 2000-2026`.
@@ -515,19 +516,51 @@ interpret them.
    number, a run date.
 5. Sentence case. No trailing full stop.
 
+**Primary category — required before recipe registration**
+Run `mr-data dataset categories --json` to read the fixed vocabulary and its definitions,
+exclusions and examples. Choose exactly one primary category by interpreting the dataset purpose,
+tables and measured entities. Save it with `mr-data dataset set DATASET_ID --category CATEGORY_ID`
+and verify `category_id` in `dataset show`. Put this dataset’s stable ID in
+`dataset.id` in the recipe file (the HTTP command wraps it under `document`) when registering; enforced builds cannot bind by name alone. The updated recipe client requests server enforcement
+of this saved category; it never guesses from tags. Draft pages may remain unassigned during research.
+
+The allowed IDs and labels are:
+
+| ID | Category |
+| --- | --- |
+| `climate-environment` | Climate & Environment |
+| `crypto` | Crypto |
+| `finance` | Finance |
+| `economics-business` | Economics & Business |
+| `government-politics` | Government & Politics |
+| `people-society` | People & Society |
+| `health-medicine` | Health & Medicine |
+| `science-research` | Science & Research |
+| `technology` | Technology |
+| `energy` | Energy |
+| `agriculture-food` | Agriculture & Food |
+| `transport-mobility` | Transport & Mobility |
+| `places-infrastructure` | Places & Infrastructure |
+| `education` | Education |
+| `sports` | Sports |
+| `arts-media-entertainment` | Arts, Media & Entertainment |
+| `law-public-safety` | Law & Public Safety |
+| `other` | Other |
+
+Classify the main subject, not the source, cadence, file format or secondary variables. Bitcoin
+price markets are Crypto; airport weather observations are Climate & Environment; crop yields
+with weather covariates are Agriculture & Food; Steam player counts are Arts, Media & Entertainment.
+For mixed datasets, use their principal analytical purpose. If none fits, choose Other and record
+why the listed definitions exclude the subject. Uncertainty alone is not a reason to use Other.
+Preserve the saved category on routine refreshes and recipe revisions. Reconsider it only when
+the dataset's principal subject changes. Never invent, combine or subdivide category IDs.
+
 **Topic rules**
-1. 3 to 8 topics, comma-separated, no duplicates.
-2. Topic one is the category, chosen from this list and no other:
-   `weather, climate, air quality, hydrology, energy, transport, aviation, public health,
-   economics, finance, elections, sports, prediction markets, demographics, geospatial,
-   agriculture`.
-3. Then, in this order, one each where it applies: the city or region, the country, the
-   phenomenon or instrument, the publisher or programme. After those, up to three more, each a
-   distinct thing a person would type that none of the earlier topics already names.
-4. Form: lowercase; ASCII letters, digits, spaces and hyphens only; one to three words; at most
-   40 characters.
-5. Never: a column name, a unit, a bare year, a second word from the category list, a market or
-   venue name unless the dataset is about that market.
+1. Write 3 to 8 useful descriptive tags, comma-separated, without duplicates.
+2. Tags are search keywords, not categories. No tag position determines the primary category.
+3. Use relevant subject, place, phenomenon, instrument or programme names.
+4. Use lowercase ASCII letters, digits, spaces and hyphens; at most 40 characters per tag.
+5. Omit column names, units, bare years and redundant terms.
 
 **The dataset description is the documentation, and it is read by people and by search engines.**
 `dataset.description` is Markdown, rendered on the dataset page with headings and lists, and it
@@ -592,9 +625,8 @@ choice. Give `table.description` the grain and the window.
 - title: at most 60 characters, contains the place, contains no colon, pipe or exclamation mark,
   no brand, no "dataset" or "data", no version number, no run date
 - opening paragraph: 150 to 240 characters, first full stop before character 200, no Markdown
-- topics: 3 to 8, no duplicates, first one is on the category list, all lowercase, each at most
-  40 characters, none is a column name, a unit, a year, a second word from the category list, or
-  a market or venue name unless the dataset is about that market
+- category: exactly one fixed ID saved and verified before recipe registration
+- topics: 3 to 8 descriptive tags, no duplicates, lowercase, at most 40 characters each
 - licence: an SPDX identifier the sources actually grant, or `--license` left off
 - every `##` heading in DESCRIPTION.md is one of the five, in that relative order, none repeated
 

mostlyright_data-0.17.7/src/mostlyright/data_harness/thin/categories.py

@@ -0,0 +1,9 @@
+"""Closed primary-category vocabulary mirrored from Studio's versioned contract."""
+
+import json
+from pathlib import Path
+
+TAXONOMY = json.loads(
+    Path(__file__).with_name("dataset-categories-v1.json").read_text(encoding="utf-8")
+)
+CATEGORY_IDS = tuple(entry["id"] for entry in TAXONOMY["categories"])

mostlyright_data-0.17.7/src/mostlyright/data_harness/thin/dataset-categories-v1.json

@@ -0,0 +1,132 @@
+{
+  "version": 1,
+  "assignment_policy": "Exactly one primary category, chosen from the dataset purpose, tables and measured entities. Prefer the principal subject over its source, cadence, format or secondary variables. Preserve the assignment on refresh. Other requires a recorded reason.",
+  "categories": [
+    {
+      "id": "climate-environment",
+      "label": "Climate & Environment",
+      "definition": "Weather, climate, pollution, ecosystems and environmental measurements.",
+      "exclusions": "Agricultural output belongs to Agriculture & Food; energy supply belongs to Energy.",
+      "examples": "Hourly weather observations; air quality; ocean temperatures"
+    },
+    {
+      "id": "crypto",
+      "label": "Crypto",
+      "definition": "Cryptocurrencies, blockchain networks, DeFi and digital-asset markets.",
+      "exclusions": "Use Finance for conventional securities, banking and non-crypto prediction markets.",
+      "examples": "Bitcoin prices; Ethereum transactions; DeFi lending"
+    },
+    {
+      "id": "finance",
+      "label": "Finance",
+      "definition": "Banking, securities, investment, insurance and financial-market activity.",
+      "exclusions": "Crypto assets belong to Crypto; economy-wide indicators belong to Economics & Business.",
+      "examples": "Stock prices; interest-rate products; prediction-market order books"
+    },
+    {
+      "id": "economics-business",
+      "label": "Economics & Business",
+      "definition": "Economic indicators, businesses, commerce, industry and consumer activity.",
+      "exclusions": "Financial instruments belong to Finance; individual population characteristics belong to People & Society.",
+      "examples": "GDP; inflation; company directories; retail sales"
+    },
+    {
+      "id": "government-politics",
+      "label": "Government & Politics",
+      "definition": "Elections, public policy, government institutions and administration.",
+      "exclusions": "Courts and crime belong to Law & Public Safety; economic outcomes belong to Economics & Business.",
+      "examples": "Election results; legislative votes; public budgets"
+    },
+    {
+      "id": "people-society",
+      "label": "People & Society",
+      "definition": "Population, demographics, communities, social conditions and human behavior.",
+      "exclusions": "Clinical outcomes belong to Health & Medicine; formal learning belongs to Education.",
+      "examples": "Census; migration; public opinion; household surveys"
+    },
+    {
+      "id": "health-medicine",
+      "label": "Health & Medicine",
+      "definition": "Clinical medicine, healthcare, diseases and public health.",
+      "exclusions": "Non-health scientific experiments belong to Science & Research.",
+      "examples": "Disease surveillance; hospitals; clinical trials"
+    },
+    {
+      "id": "science-research",
+      "label": "Science & Research",
+      "definition": "Physical and natural sciences, experiments, astronomy and general research metadata.",
+      "exclusions": "Use the subject-specific category for weather, health, technology or other covered research.",
+      "examples": "Astronomical observations; particle experiments; research publications"
+    },
+    {
+      "id": "technology",
+      "label": "Technology",
+      "definition": "Software, computing, AI, internet and telecommunications.",
+      "exclusions": "Blockchain belongs to Crypto; media consumption and gaming belong to Arts, Media & Entertainment.",
+      "examples": "Software packages; internet performance; AI benchmarks"
+    },
+    {
+      "id": "energy",
+      "label": "Energy",
+      "definition": "Electricity, fuels, energy production, consumption and supply infrastructure.",
+      "exclusions": "Weather measurements belong to Climate & Environment even when used for energy analysis.",
+      "examples": "Power generation; electricity demand; oil production"
+    },
+    {
+      "id": "agriculture-food",
+      "label": "Agriculture & Food",
+      "definition": "Farming, livestock, fisheries, food production and food systems.",
+      "exclusions": "Weather alone belongs to Climate & Environment; clinical nutrition outcomes belong to Health & Medicine.",
+      "examples": "Crop yields with weather covariates; livestock counts; fisheries"
+    },
+    {
+      "id": "transport-mobility",
+      "label": "Transport & Mobility",
+      "definition": "Movement of people and goods by road, rail, air and sea.",
+      "exclusions": "Airport weather observations belong to Climate & Environment.",
+      "examples": "Transit ridership; flight movements; shipping routes"
+    },
+    {
+      "id": "places-infrastructure",
+      "label": "Places & Infrastructure",
+      "definition": "Geography, maps, property, buildings, housing and non-energy infrastructure.",
+      "exclusions": "Movement belongs to Transport & Mobility; electricity systems belong to Energy.",
+      "examples": "Building footprints; property listings; water networks"
+    },
+    {
+      "id": "education",
+      "label": "Education",
+      "definition": "Schools, universities, learning, students and educational outcomes.",
+      "exclusions": "Research datasets use their subject category rather than the institution that produced them.",
+      "examples": "School enrollment; exam results; course catalogs"
+    },
+    {
+      "id": "sports",
+      "label": "Sports",
+      "definition": "Sports competitions, athletes, teams and athletic performance.",
+      "exclusions": "Video-game activity and esports belong to Arts, Media & Entertainment.",
+      "examples": "Match results; player statistics; athletic records"
+    },
+    {
+      "id": "arts-media-entertainment",
+      "label": "Arts, Media & Entertainment",
+      "definition": "Culture, arts, news media, music, film, gaming, esports and audience activity.",
+      "exclusions": "A news dataset about a specific subject belongs to that subject when the observations measure that subject.",
+      "examples": "Film releases; Steam players; game-related pageviews; music charts"
+    },
+    {
+      "id": "law-public-safety",
+      "label": "Law & Public Safety",
+      "definition": "Legal systems, courts, crime, policing and emergency response.",
+      "exclusions": "Election and legislative activity belongs to Government & Politics.",
+      "examples": "Court decisions; crime reports; emergency incidents"
+    },
+    {
+      "id": "other",
+      "label": "Other",
+      "definition": "A meaningful subject not covered by any category after reviewing its definitions.",
+      "exclusions": "Never use for missing research, uncertainty, mixed subjects, test status, cadence or file format. Record why none fits.",
+      "examples": "A documented subject outside the defined taxonomy"
+    }
+  ]
+}

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/src/mostlyright/data_harness/thin/recipe.py

@@ -323,11 +323,9 @@ def column_description_warnings(absent: list[str]) -> list[str]:
 def check_dataset_binding(dataset: Any) -> None:
     """Refuse a ``dataset.id`` that could not name a dataset, before anything is sent.
 
-    ⚠ THE MEMBER IS OPTIONAL, AND ITS ABSENCE IS NOT A MISTAKE. A document with no ``dataset.id``
-    registers exactly as it always did: Studio resolves the container from ``(workspace_id, name)``.
-    What the member buys is the live-authoring order -- ``mr-data dataset create`` mints the page
-    BEFORE the research starts, so somebody watching sees it fill in, and the registration binds to
-    the page they are already looking at instead of minting a second one behind them.
+    The legacy server contract admits name-only registration, but this CLI requests category
+    enforcement, which requires a saved category and stable dataset.id. The build creates the
+    page before research and binds the recipe to that same page. Studio checks the saved category.
 
     ⚠ AND A MISTYPED ONE IS THE EXPENSIVE MISTAKE, WHICH IS WHY IT IS CHECKED HERE. An identifier
     that is not a UUID is a 422 after a token exchange; an identifier that is a UUID and names
@@ -354,8 +352,7 @@ def check_dataset_binding(dataset: Any) -> None:
         raise ThinLaneError(
             "THIN_RECIPE_DOCUMENT_INVALID",
             "the recipe document's dataset.id is not a dataset identifier; it is the identifier "
-            "mr-data dataset create printed, or the member is left out and Studio resolves the "
-            "dataset from its name",
+            "mr-data dataset create printed. Save its category before registering the recipe",
         ) from None
     if stated != canonical:
         raise ThinLaneError(

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/src/mostlyright/data_harness/thin/v4.py

@@ -661,7 +661,11 @@ class StudioV4Client(StudioApiClient):
         return self._call(
             "POST",
             CREATE_RECIPE_PATH,
-            body={"workspace_id": str(self.session.workspace_id), "document": dict(document)},
+            body={
+                "workspace_id": str(self.session.workspace_id),
+                "document": dict(document),
+                "require_category": True,
+            },
             expected=(201,),
         )
 

{mostlyright_data-0.17.6 → mostlyright_data-0.17.7}/src/mostlyright/data_harness/thin/v4_datasets.py

@@ -107,6 +107,7 @@ from mostlyright.data_harness.thin.activity import (
 from mostlyright.data_harness.thin.activity import (
     report as _report_activity,
 )
+from mostlyright.data_harness.thin.categories import CATEGORY_IDS, TAXONOMY
 from mostlyright.data_harness.thin.narrative import (
     NOT_EVIDENCE,
     cell_id_from_heading,
@@ -187,6 +188,7 @@ COMMAND_HELP = "create the dataset page, correct it, write its record, share it,
 #: and version stays where it is. That is why the confirmation is the dataset's own name typed out
 #: rather than a yes: a person who has the wrong page open cannot type the right name.
 ACTIONS: tuple[str, ...] = (
+    "categories",
     "create",
     "set",
     "draft",
@@ -246,6 +248,7 @@ DATASET_MEMBERS: tuple[str, ...] = (
     # read back is a write an agent cannot check.
     "topics",
     "license",
+    "category_id",
     "status",
     "version",
     "cover",
@@ -602,10 +605,16 @@ def declare_arguments(parser: argparse.ArgumentParser) -> None:
             "a description is paragraphs and argv is one line"
         ),
     )
+    parser.add_argument(
+        "--category",
+        choices=CATEGORY_IDS,
+        help="with set, the primary category; inspect definitions with dataset categories",
+    )
     parser.add_argument(
         "--topics",
         help=(
-            "with set, the categories this dataset is filed under, as one comma-separated list. "
+            "with set, optional descriptive tags, separate from the primary category, "
+            "as one comma-separated list. "
             "Each entry is trimmed and lowercased here and checked by Studio, which holds the "
             "bounds: at most eight, no duplicates, and each one 1 to 40 characters of lowercase "
             "letters, digits, spaces and hyphens. An empty list -- --topics '' -- takes them off "
@@ -1161,15 +1170,29 @@ def write(
     named = getattr(args, "name", None)
     name = checked_name(named, needed_by="set") if named is not None else None
     description = read_description(args, stdin=stdin)
+    category = getattr(args, "category", None)
+    if category is not None and category not in CATEGORY_IDS:
+        raise ThinLaneError(
+            "THIN_REQUEST_INVALID", "--category must be one of: " + ", ".join(CATEGORY_IDS)
+        )
     topics = read_topics(getattr(args, "topics", None))
     named_license = getattr(args, "license", None)
-    if name is None and description is None and topics is None and named_license is None:
+    if (
+        name is None
+        and description is None
+        and topics is None
+        and named_license is None
+        and category is None
+    ):
         raise ThinLaneError(
             "THIN_REQUEST_INVALID",
             "mr-data dataset set writes the title, the topics, the licence and the description: "
-            "give --name, --topics, --license, --description-file, or any of them together",
+            "give --name, --category, --topics, --license, --description-file, "
+            "or any of them together",
         )
     fields: dict[str, Any] = {}
+    if category is not None:
+        fields["category_id"] = category
     if name is not None:
         fields["name"] = name
     if description is not None:
@@ -1865,6 +1888,8 @@ def dataset(
     # would drop them silently. They belong to `set` alone: `create` mints a page from a working
     # title and `draft` writes a snapshot of text, and neither has anywhere to put a category or
     # a licence.
+    if action != "set" and getattr(args, "category", None) is not None:
+        raise ThinLaneError("THIN_REQUEST_INVALID", "--category belongs to mr-data dataset set")
     if action != "set" and (
         getattr(args, "topics", None) is not None or getattr(args, "license", None) is not None
     ):
@@ -1873,6 +1898,9 @@ def dataset(
             f"--topics and --license belong to mr-data dataset set, not to {action}",
         )
 
+    if action == "categories":
+        return {"schema_version": "mostlyright-dataset-categories.v1", **TAXONOMY}
+
     if action in _NEEDS_DATASET and not getattr(args, "dataset_id", None):
         raise ThinLaneError(
             "THIN_REQUEST_INVALID",