Parameters

This reference maps Earthdata MCP tool parameters to the upstream API arguments and schema fields behind them. It provides consumers with a clear picture of current API integration depth and search capabilities. Each tool’s section names the endpoint and schema it is built on.

All search tools globally support the limit, cursor, and fields parameters for pagination and response filtering. These are omitted from the tables below for brevity.

Note on Harmony tools: Unlike the CMR-backed tools above, the four Harmony transformation tools below do not query CMR’s /search endpoints or use UMM schemas. They call the Harmony API directly via the official harmony-py client library. Parameter/field tables for these tools reference harmony-py and the Harmony API’s own JSON response shapes instead of CMR API parameters and UMM JSON paths.

Authentication required: Unlike the CMR-backed discovery tools, all four Harmony tools require an authenticated Earthdata Login access token (retrieved via get_access_token()) to call Harmony’s API on the user’s behalf.

`get_collections`

Searches for datasets (collections) using scientific keywords, instruments, platforms, or spatial/temporal constraints.

CMR Endpoint: /search/collections
Schema: UMM-C (v1.18.3)

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`keyword`	`keyword`	Free-text search across a collection’s indexed metadata, case insensitive. All words must match, so extra terms narrow the search rather than widen it.
✅	`concept_id`	`concept_id`	Exact CMR concept ID, e.g. `C2036882064-POCLOUD`. Direct lookup of a known collection.
✅	`short_name`	`short_name`	Collection short name, e.g. `MOD11A1`. Exact match unless the `*` or `?` wildcards are used.
✅	`provider`	`provider`	Archive center that published the collection, e.g. `PODAAC`, `GES_DISC`. Cloud migration has renamed some providers, and a stale ID returns no results rather than an error.
✅	`temporal_start_date`	`temporal`	Start of the temporal window, ISO 8601. Matches collections whose declared range overlaps it.
✅	`temporal_end_date`	`temporal`	End of the temporal window, ISO 8601. Matches collections whose declared range overlaps it.
✅	`spatial_wkt_geometry`	`polygon, point, bounding_box`	Area of interest as a WKT `POLYGON`, `POINT`, or `LINESTRING`. Matches collections whose declared extent intersects it, so a precise shape avoids false positives.
✅	`platform`	`platform`	Platform short names to filter by, e.g. `Terra`, `Aqua`. Narrows a search to one mission’s instruments.
✅	`instrument`	`instrument`	Instrument short names to filter by, e.g. `MODIS`, `VIIRS`. Matches the instrument field rather than the whole record, so it is tighter than the same name as a keyword.
✅	`processing_level_id`	`processing_level_id`	Processing levels to filter by, e.g. `3`, `3A`. Distinguishes L2 swath from L3 gridded products.
✅	`has_granules`	`has_granules`	When true, limits results to collections that hold granules. CMR also carries metadata-only records for planned missions and datasets archived elsewhere.
❌	N/A	`doi`	Search by digital object identifier
❌	N/A	`project`	Search by project/campaign name
❌	N/A	`data_center`	Search by data center/archive center
❌	N/A	`science_keywords`	Search by GCMD science keywords hierarchy
❌	N/A	`updated_since`	Filter by recently updated collections

Output Fields

Status	MCP Response Field	UMM JSON Path	Transformed	Description
✅	`abstract`	`Abstract`		Collection summary or abstract
✅	`archive_and_distribution_information`	`ArchiveAndDistributionInformation`	✅	File formats and media types (e.g., [{format, media_type}])
✅	`bounding_box`	`SpatialExtent.HorizontalSpatialDomain.Geometry.BoundingRectangles`	✅	[West, South, East, North] Minimum Bounding Rectangle
✅	`collection_data_type`	`CollectionDataType`		e.g., SCIENCE_QUALITY, NEAR_REAL_TIME
✅	`collection_progress`	`CollectionProgress`		ACTIVE, COMPLETE, DEPRECATED, or PLANNED
✅	`concept_id`	`meta.concept-id`		CMR collection concept ID
✅	`data_centers`	`DataCenters`	✅	Archiving DAACs — array of {role, short_name}
✅	`doi`	`DOI`		Digital Object Identifier
✅	`entry_title`	`EntryTitle`		Collection title
✅	`instruments`	`Platforms.Instruments.ShortName`	✅	Instrument short names
✅	`is_ongoing`	`IsOngoing`	✅	Whether the collection is ongoing
✅	`native_id`	`meta.native-id`		The native ID of the collection record
✅	`platforms`	`Platforms.ShortName`	✅	Platform short names
✅	`processing_level_id`	`ProcessingLevel.Id`		Processing level (e.g., L3, L4)
✅	`provider_id`	`meta.provider-id`		The provider ID of the collection
✅	`related_urls`	`RelatedUrls`	✅	List of related URLs (e.g., documentation, guides)
✅	`revision_id`	`meta.revision-id`		The revision ID of the collection metadata
✅	`science_keywords`	`ScienceKeywords`		GCMD science keyword hierarchy (Category/Topic/Term/VariableLevel)
✅	`short_name`	`ShortName`		Collection short name
✅	`spatial_resolution`	`SpatialExtent.HorizontalSpatialDomain.ResolutionAndCoordinateSystem`	✅	Human-readable spatial resolution
✅	`temporal_resolution`	`TemporalExtents.TemporalResolution`	✅	Human-readable temporal resolution
✅	`time_end`	`TemporalExtents.RangeDateTimes.EndingDateTime`	✅	End of temporal coverage
✅	`time_start`	`TemporalExtents.RangeDateTimes.BeginningDateTime`	✅	Start of temporal coverage
✅	`version`	`Version`		Collection version
❌	N/A	`AccessConstraints`		Access constraints and authorization requirements
❌	N/A	`AdditionalAttributes`		Provider-specific additional attributes
❌	N/A	`ContactGroups/ContactPersons`		Point of contact information
❌	N/A	`Projects`		Projects or campaigns associated with the collection
❌	N/A	`SpatialKeywords`		Geographic location keywords

`get_granules`

Searches for specific data files (granules) within a collection to verify actual data availability for a given time and location.

CMR Endpoint: /search/granules
Schema: UMM-G (v1.6.5)

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`collection_concept_id`	`collection_concept_id`	Parent collection to search within, e.g. `C2723758340-GES_DISC`. Required.
✅	`temporal_start_date`	`temporal`	Start of the temporal window, ISO 8601. Matches granules whose coverage overlaps it.
✅	`temporal_end_date`	`temporal`	End of the temporal window, ISO 8601. Matches granules whose coverage overlaps it.
✅	`spatial_wkt_geometry`	`polygon, point, bounding_box`	Area of interest as a WKT `POLYGON`, `POINT`, or `LINESTRING`. Matches granules whose footprint intersects it, so a precise shape avoids false positives.
✅	`cloud_cover_min`	`cloud_cover`	Lower bound on cloud cover, 0 to 100. Optical imagery only; SAR and altimetry do not report it.
✅	`cloud_cover_max`	`cloud_cover`	Upper bound on cloud cover, 0 to 100. Use `20` or lower for mostly clear scenes. Optical imagery only.
✅	`day_night_flag`	`day_night_flag`	Acquisition lighting, one of `DAY`, `NIGHT`, or `UNSPECIFIED`.
✅	`sort_key`	`sort_key`	Result ordering, e.g. `-start_date` for newest first. Defaults to relevance, not recency, so recent data is not returned first unless asked for.
❌	N/A	`granule_ur`	Search by exact granule UR
❌	N/A	`producer_granule_id`	Search by producer granule ID
❌	N/A	`readable_granule_name`	Search by either granule UR or producer ID
❌	N/A	`orbit_number`	Filter granules by orbit number
❌	N/A	`updated_since`	Filter by recently updated granules

Output Fields

Status	MCP Response Field	UMM JSON Path	Transformed	Description
✅	`access_urls`	`RelatedUrls`	✅	Actionable data access URLs (Note: Access requires Earthdata Login authentication)
✅	`additional_attributes`	`AdditionalAttributes`	✅	Provider-specific attributes (e.g., tile coords, quality flags) — array of {name, values[]}
✅	`bounding_box`	`SpatialExtent.HorizontalSpatialDomain.Geometry.BoundingRectangles`	✅	`[West, South, East, North]` minimum bounding rectangle. Encloses swath and irregular footprints, so the corners may hold no data.
✅	`cloud_cover`	`CloudCover`		Cloud cover percentage
✅	`collection_concept_id`	`CollectionReference.ShortName/Version`		Parent collection concept ID
✅	`concept_id`	`meta.concept-id`		CMR granule concept ID
✅	`data_format`	`DataFormat`	✅	File format (e.g., NetCDF-4, GeoTIFF)
✅	`day_night_flag`	`DataGranule.DayNightFlag`		DAY, NIGHT, BOTH, or UNSPECIFIED
✅	`granule_ur`	`GranuleUR`		Granule UR
✅	`native_id`	`meta.native-id`		The native ID of the granule record
✅	`orbit_info`	`SpatialExtent.OrbitCalculatedSpatialDomains`	✅	Orbit calculated spatial domains — array of {orbit_number, equator_crossing_longitude, equator_crossing_date_time}
✅	`producer_granule_id`	`DataGranule.ProducerGranuleId`		Producer granule ID
✅	`production_date`	`DataGranule.ProductionDateTime`	✅	Date the granule was generated (ProductionDateTime)
✅	`provider_id`	`meta.provider-id`		The provider ID of the granule
✅	`revision_id`	`meta.revision-id`		The revision ID of the granule metadata
✅	`size_mb`	`DataGranule.ArchiveAndDistributionInformation`	✅	Size of the data granule in MB
✅	`time_end`	`TemporalExtent.RangeDateTime.EndingDateTime`	✅	Granule temporal end
✅	`time_start`	`TemporalExtent.RangeDateTime.BeginningDateTime`	✅	Granule temporal start
❌	N/A	`AccessConstraints`		Access constraints and authorization requirements
❌	N/A	`InputGranules`		Provenance information about source granules
❌	N/A	`MeasuredParameters`		Variables/parameters measured in the granule
❌	N/A	`Platforms`		Specific platforms used for the granule
❌	N/A	`Projects`		Projects or campaigns associated with the granule

`get_variables`

Discovers scientific variables and measurements associated with a collection, or looks up variables by keyword.

CMR Endpoint: /search/variables
Schema: UMM-V (v1.9.0)

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`collection_concept_id`	`concept_id`	Collection to find variables for, e.g. `C2763266360-LPCLOUD`.
✅	`keyword`	`keyword`	Free-text keyword to discover variables without a collection ID.
❌	N/A	`name`	Exact match on variable name
❌	N/A	`provider`	Filter by provider ID

Output Fields

Status	MCP Response Field	UMM JSON Path	Transformed	Description
✅	`concept_id`	`meta.concept-id`		CMR variable concept ID
✅	`name`	`Name`		The short name of the variable
✅	`long_name`	`LongName`		The long name of the variable
✅	`definition`	`Definition`		The definition of the variable
✅	`data_type`	`DataType`		The data type of the variable
✅	`units`	`Units`		The units of the variable
✅	`scale`	`Scale`		The scale factor for the variable data
✅	`offset`	`Offset`		The offset for the variable data
✅	`fill_values`	`FillValues`		Fill values used for missing or invalid data
✅	`valid_ranges`	`ValidRanges`		Valid data ranges for the variable
✅	`dimensions`	`Dimensions`		Dimensions associated with the variable
✅	`standard_name`	`StandardName`		The CF Standard Name of the variable
✅	`science_keywords`	`ScienceKeywords`		GCMD Science Keywords hierarchy
✅	`variable_type`	`VariableType`		Type of variable (e.g., SCIENCE_VARIABLE, COORDINATE)
✅	`variable_sub_type`	`VariableSubType`		Sub-type of variable
✅	`sets`	`Sets`		Logical groupings for the variable
✅	`measurement_identifiers`	`MeasurementIdentifiers`		Measurement context and provenance
✅	`sampling_identifiers`	`SamplingIdentifiers`		Sampling method context
✅	`related_urls`	`RelatedUrls`	✅	URLs specific to the variable
❌	N/A	`AdditionalIdentifiers`		Additional identifiers for the variable
❌	N/A	`IndexRanges`		Array index ranges for the variable
❌	N/A	`InstanceInformation`		Variable instance information

`get_tools`

Finds web portals and downloadable software associated with a collection, returning URLs and deep-linking templates.

CMR Endpoint: /search/tools
Schema: UMM-T (v1.2.0)

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`collection_concept_id`	`concept_id`	Collection to find associated tools for, e.g. `C2723758340-GES_DISC`.
✅	`keyword`	`keyword`	Free-text keyword to discover tools without a collection ID.
❌	N/A	`name`	Exact match on tool name
❌	N/A	`provider`	Filter by provider ID

Output Fields

Status	MCP Response Field	UMM JSON Path	Transformed	Description
✅	`access_constraints`	`AccessConstraints`		Constraints for accessing the tool
✅	`concept_id`	`meta.concept-id`		CMR tool concept ID
✅	`description`	`Description`		A brief description of the tool
✅	`doi`	`DOI`		Digital Object Identifier of the tool
✅	`long_name`	`LongName`		The long name of the tool
✅	`name`	`Name`		The name of the tool
✅	`native_id`	`meta.native-id`		The native ID of the tool record
✅	`organizations`	`Organizations`		Organizations responsible for the tool
✅	`potential_action`	`PotentialAction`		Smart handoff definition for parameterized deep links
✅	`provider_id`	`meta.provider-id`		The provider ID of the tool
✅	`quality`	`Quality`		Quality information about the tool
✅	`related_urls`	`RelatedUrls`	✅	Documentation, guides, or other related links
✅	`revision_id`	`meta.revision-id`		The revision ID of the tool metadata
✅	`supported_browsers`	`SupportedBrowsers`		Browsers and versions supported by the tool
✅	`supported_input_formats`	`SupportedInputFormats`		List of input format names supported by the tool
✅	`supported_operating_systems`	`SupportedOperatingSystems`		Operating systems and versions supported by the tool
✅	`supported_output_formats`	`SupportedOutputFormats`		List of output format names supported by the tool
✅	`supported_software_languages`	`SupportedSoftwareLanguages`		Programming languages and versions supported by the tool
✅	`tool_keywords`	`ToolKeywords`		Earth science keywords representative of the tool
✅	`type`	`Type`		The type of the tool (e.g., Downloadable Tool, Web User Interface, Web Portal, Model)
✅	`url`	`URL`		Primary URL for accessing the tool
✅	`use_constraints`	`UseConstraints`		Restrictions or limitations on using the tool
✅	`version`	`Version`		The edition or version of the tool
❌	N/A	`AncillaryKeywords`		Additional keywords for the tool
❌	N/A	`ContactGroups/ContactPersons`		Point of contact information
❌	N/A	`LastUpdatedDate`		When the tool metadata was last updated

`get_services`

Discovers data access endpoints and visualization layers associated with a collection.

CMR Endpoint: /search/services
Schema: UMM-S (v1.5.3)

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`collection_concept_id`	`concept_id`	Collection to find associated services for, e.g. `C2723758340-GES_DISC`.
✅	`keyword`	`keyword`	Free-text keyword to discover services without a collection ID.
✅	`type`	`type`	Service type to filter by, e.g. `OPeNDAP`, `WMS`, `Harmony`.
❌	N/A	`name`	Exact match on service name
❌	N/A	`provider`	Filter by provider ID

Output Fields

Status	MCP Response Field	UMM JSON Path	Transformed	Description
✅	`access_constraints`	`AccessConstraints`		Authentication or authorization requirements
✅	`concept_id`	`meta.concept-id`		CMR service concept ID
✅	`description`	`Description`		A brief description of the service
✅	`long_name`	`LongName`		The long name of the service
✅	`name`	`Name`		The name of the service
✅	`native_id`	`meta.native-id`		The native ID of the service record
✅	`operation_metadata`	`OperationMetadata`		Operation names and distributed computing platform
✅	`provider_id`	`meta.provider-id`		The provider ID of the service
✅	`related_urls`	`RelatedUrls`	✅	Documentation, guides, or other related links
✅	`revision_id`	`meta.revision-id`		The revision ID of the service metadata
✅	`service_keywords`	`ServiceKeywords`		Controlled vocabulary for service capability
✅	`service_options`	`ServiceOptions`		Subset types, supported projections, output formats
✅	`service_organizations`	`ServiceOrganizations`	✅	Organizations that run the service endpoint
✅	`type`	`Type`		The type of the service
✅	`url`	`URL`		Primary endpoint URL information
✅	`use_constraints`	`UseConstraints`		Legal restrictions or usage limits
✅	`version`	`Version`		The edition or version of the service
❌	N/A	`AncillaryKeywords`		Additional keywords for the service
❌	N/A	`ContactGroups/ContactPersons`		Point of contact information
❌	N/A	`ServiceQuality`		Information about service quality

`get_keywords`

Discovers official Earthdata scientific vocabulary terms to translate colloquial user inputs into precise search labels.

CMR Endpoint: /search/keywords
Schema: KMS Concept (v2.0)

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`query`	`pattern`	Term to search for across KMS schemes, e.g. `moisture`.
✅	`scheme`	`keyword_scheme`	A single KMS scheme to search, e.g. `sciencekeywords`, `platforms`, `instruments`. Searches every scheme if omitted. Full list.

Output Fields

Status	MCP Response Field	UMM JSON Path	Description
✅	`uuid`	`uuid`	The unique UUID of the KMS concept
✅	`prefLabel`	`prefLabel`	The preferred label of the KMS concept
✅	`scheme`	`scheme`	The scheme the concept belongs to
✅	`definition`	`definition`	The primary definition of the concept, if available

`get_citations`

Discovers citation records (publications, DOIs) associated with a collection, or looks up a citation directly by identifier.

CMR Endpoint: /search/citations
Schema: Citation (v1.0.0), a CMR generic document type. Citations have not been adopted into UMM yet, so there is no published schema to link.

Exactly one of collection_concept_id or identifier is required. Supplying both, or neither, is a validation error.

collection_concept_id is not a CMR citation search parameter. The tool first reads the collection record’s citation associations, then searches citations by the concept IDs it found, so a collection with no associations returns no results without a second request.

Input Parameters

Status	MCP Argument	CMR API Parameter	Description
✅	`collection_concept_id`	`concept_id[]`	Collection to find citations for, e.g. `C2763266360-LPCLOUD`. Resolved through the collection’s citation associations.
✅	`identifier`	`identifier`	DOI or other citation identifier, e.g. `10.24193/AWC2022_05`. Looks up a single citation directly.
✅	`provider`	`provider`	CMR provider that published the citation, e.g. `ESDIS`. Combines with a collection or identifier lookup.
❌	N/A	`name`	Search by citation name. Exact match, not free text.
❌	N/A	`title`	Search by title. Exact match, not free text.
❌	N/A	`year`	Filter by publication year
❌	N/A	`author-name`	Search by author name
❌	N/A	`author-orcid`	Search by author ORCID
❌	N/A	`container`	Search by container or journal name
❌	N/A	`type`	Search by citation type (e.g., journal-article, proceedings-article)
❌	N/A	`identifier-type`	Filter by identifier type (e.g., DOI)
❌	N/A	`resolution-authority`	Filter by the authority that resolves the identifier
❌	N/A	`relationship-type`	Filter by relationship to the cited work (e.g., Cites, Refers)
❌	N/A	`related-identifier`	Search by an identifier the citation relates to
❌	N/A	`related-identifier-with-type`	Search relationship and identifier as a pair (e.g., `Cites:10.5067/SAMPLE/DATA`)
❌	N/A	`concept-id`	Direct lookup by citation concept ID
❌	N/A	`native-id`	Lookup by native ID
❌	N/A	`id`	Lookup by citation ID
❌	N/A	`keyword`	Search by science keyword. Accepted by CMR but returns no matches for citation records.

Output Fields

Every field in the Citation schema is surfaced.

Status	MCP Response Field	Record JSON Path	Transformed	Description
✅	`abstract`	`Abstract`		Abstract or description of the cited work
✅	`associated_collections`	`meta.associations.collections`	✅	Concept IDs of collections this citation is associated with. Pass these to `get_collections` for the dataset details.
✅	`citation_metadata`	`CitationMetadata`		Nested bibliographic metadata (Author, Year, Publisher, Title, Container, Volume, Pages, and related fields)
✅	`concept_id`	`meta.concept-id`		CMR citation concept ID
✅	`identifier`	`Identifier`		The primary identifier, usually a DOI
✅	`identifier_type`	`IdentifierType`		Type of the identifier (e.g., DOI)
✅	`metadata_specification`	`MetadataSpecification`		Schema name, version, and URL for the citation record
✅	`name`	`Name`		The name or title of the citation
✅	`native_id`	`meta.native-id`		The native ID of the citation record
✅	`provider_id`	`meta.provider-id`		The provider ID of the citation
✅	`related_identifiers`	`RelatedIdentifiers`		Related works and datasets, each with a relationship type (e.g., Cites, Refers)
✅	`resolution_authority`	`ResolutionAuthority`		Authority that resolves the identifier (e.g., https://doi.org)
✅	`revision_id`	`meta.revision-id`		The revision ID of the citation metadata

`get_transformation_options`

Looks up which Harmony operations a collection supports (subsetting, reprojection, averaging, concatenation), which services implement them, supported output formats, and variables.

Underlying API: Harmony /capabilities endpoint
Response Format: Capabilities response version 3 (pinned server-side; not exposed as a tool input)

Input Parameters

Status	MCP Argument	Harmony API Parameter	Description
✅	`collection_concept_id`	`collectionId`	Concept ID of the collection (format: C<number>-<PROVIDER>). Exactly one of `collection_id` or `short_name` must be provided.
✅	`short_name`	`shortName`	Short name of the collection. Exactly one of `collection_id` or `short_name` must be provided. If multiple collections share the short name, Harmony prefers the one configured for use in Harmony.
❌	N/A	`version`	Capabilities response format version — pinned to ‘3’ server-side, not user-configurable

Output Fields

Status	MCP Response Field	Harmony Capabilities Field	Transformed	Description
✅	`concept_id`	`conceptId`		Concept ID of the collection
✅	`short_name`	`shortName`		Short name of the collection
✅	`summary`	(top-level capability fields)	✅	Collection-level summary object combining:
↳	`summary.subsetting.bbox`	`subsetting.bbox`		True if bounding box subsetting is supported
↳	`summary.subsetting.dimension`	`subsetting.dimension`		True if dimension subsetting is supported
↳	`summary.subsetting.shape`	`subsetting.shape`		True if shape (polygon) subsetting is supported
↳	`summary.subsetting.temporal`	`subsetting.temporal`		True if temporal subsetting is supported
↳	`summary.subsetting.variable`	`subsetting.variable`		True if variable subsetting is supported
↳	`summary.reprojection.supported`	`reprojection.supported`		True if reprojection is supported
↳	`summary.reprojection.supported_projections`	`reprojection.supportedProjections`	✅	Output projections supported across applicable services — each with `name` and `crs` (e.g., ‘EPSG:4326’)
↳	`summary.reprojection.interpolation_methods`	`reprojection.interpolationMethods`	✅	Interpolation methods supported for resampling during reprojection
↳	`summary.averaging.time`	`averaging.time`		True if time averaging is supported
↳	`summary.averaging.area`	`averaging.area`		True if area averaging is supported
↳	`summary.concatenation`	`concatenation`		True if concatenation is supported
↳	`summary.output_formats`	`outputFormats`	✅	Output formats supported across all applicable services — each with `name` and `mime_type`
✅	`services`	`services`	✅	Harmony services applicable to the collection, each with `name`, `href` (link to the service’s UMM-S record in CMR), and a per-service `capabilities` object (same shape as `summary`) — some capability combinations are only valid on specific services, so check here to determine which combinations a given service supports
✅	`variables`	`variables`	✅	Variables associated with the collection, each with `name`, `long_name`, `href` (link to the variable’s UMM-Var record in CMR), `units`, and `science_keywords` (GCMD Category/Topic/Term/VariableLevel1-3/DetailedVariable)
✅	`capabilities_version`	`capabilitiesVersion`		Version identifier of this capabilities response format (currently pinned to ‘3’)

`submit_transformation_job`

Submits a Harmony data processing request (subset/reformat/reproject).

Underlying Library: harmony-py — Request + Client.submit()
Underlying Endpoint: Harmony job submission API (OGC Coverages-based)

Input Parameters

Status	MCP Argument	harmony-py `Request` Parameter	Description
✅	`collection_concept_id`	`collection` (via `Collection(id=...)`)	CMR concept ID of the collection (e.g., C1234567-PROV).
✅	`bbox`	`spatial` (via `BBox(west, south, east, north)`)	Bounding box array in degrees: [west, south, east, north].
✅	`shape`	`shape`	Local path to a supported Harmony shapefile-subsetting file (.zip/.shz/.json/.geojson/.kml).
✅	`temporal_start` / `temporal_stop`	`temporal` (dict: `{'start': ..., 'stop': ...}`)	ISO 8601 start/stop timestamps for temporal subsetting.
✅	`variables`	`variables`	List of specific variable names to subset.
✅	`format`	`format`	Desired output MIME type (e.g., ‘image/tiff’, ‘application/x-netcdf4’).
✅	`crs`	`crs`	Target Coordinate Reference System for reprojection (e.g., ‘EPSG:4326’).
✅	`width`	`width`	Target width in pixels for spatial reprojection/resampling.
✅	`height`	`height`	Target height in pixels for spatial reprojection/resampling.
✅	`max_results`	`max_results`	Maximum number of granules/results to process.
✅	`granule_ids`	`granule_id`	Optional list of specific CMR granule concept IDs to process.
❌	N/A	`concatenate`	Concatenate outputs into a single file
❌	N/A	`grid`	Named output grid
❌	N/A	`extend`	Dimension(s) to extend
❌	N/A	`average`	Spatial/temporal averaging method
❌	N/A	`skip_preview`	Skip generation of preview images
❌	N/A	`destination_url`	Custom output destination (e.g., user’s own S3 bucket)
❌	N/A	`ignore_errors`	Continue job even if some granules fail
❌	N/A	`labels`	Job labels for organization/tracking

Output Fields

Same output model as get_transformation_job_status below (GetTransformationJobStatusOutput) — see that section’s Output Fields table for the full field list (job_id, status, message, progress, created_at/updated_at, links, total_hits, next_cursor, code, description, etc.). Immediately after submission, status typically reflects the job’s initial state (e.g., running or accepted) rather than a terminal state, and links will typically be empty until the job completes. code/description are populated only if the submission itself failed.

`get_transformation_job_status`

Gets the current status, progress percentage, and metadata for a Harmony job.

Underlying Library: harmony-py — Client.status()
Underlying Endpoint: Harmony GET /jobs/{jobID} API

Note: Like other search tools, this tool also supports limit, cursor, and fields — but unlike CMR-backed tools, pagination here is applied in-memory to the job’s links array after Harmony returns the full job record, rather than being passed through to an upstream API page parameter. Cursors are job-scoped and lock in the original job_id.

Input Parameters

Status	MCP Argument	harmony-py Parameter	Description
✅	`job_id`	`job_id`	The ID of the Harmony job to check.
✅	`limit`	N/A (in-memory slicing, MCP-side)	Max number of `links` to return per page (default 10).
✅	`cursor`	N/A (in-memory slicing, MCP-side)	Job-scoped pagination token over the `links` array.
✅	`fields`	N/A (MCP-side response filtering)	Restrict which output fields are returned.

Output Fields

Status	MCP Response Field	harmony-py / Job JSON Field	Transformed	Description
✅	`job_id`	`jobID`		Unique identifier (UUID) for the job
✅	`status`	`status`		Current status (`running`, `successful`, `failed`, `canceled`, etc.)
✅	`message`	`message`		Human-readable message regarding the job status
✅	`progress`	`progress`		Progress percentage of the job (0–100)
✅	`created_at`	`createdAt`		UTC timestamp (ISO 8601) when the job was created
✅	`updated_at`	`updatedAt`		UTC timestamp (ISO 8601) when the job was last updated
✅	`created_at_local`	`createdAt`	✅	Local timestamp equivalent of `created_at`
✅	`updated_at_local`	`updatedAt`	✅	Local timestamp equivalent of `updated_at`
✅	`request`	`request`		The original Harmony request URL with parameters
✅	`num_input_granules`	`numInputGranules`		Total number of input granules processed by the job
✅	`data_expiration`	`dataExpiration`		UTC timestamp (ISO 8601) when the job’s data will expire
✅	`data_expiration_local`	`dataExpiration`	✅	Local timestamp equivalent of `data_expiration`
✅	`links`	`links`	✅	Result/download links for the current page, sliced in-memory from the full job link list once terminal
✅	`total_hits`	N/A (computed from full `links` count)	✅	Full link count for the job
✅	`next_cursor`	N/A (computed, MCP-side)	✅	Opaque, job-scoped pagination token for the next page of `links`; `None` when no more results

`control_transformation_job`

Cancels, pauses, or resumes a running or paused Harmony transformation job.

Underlying Library: harmony-py — Client.cancel() / Client.pause() / Client.resume(), followed by Client.status()
Underlying Endpoint: Harmony job control API (PUT /jobs/{jobID}/cancel, /pause, /resume)

Note — valid state transitions: resume on a job that isn’t paused, or pause/cancel on a job that’s already in a terminal state (e.g., successful, failed, canceled), will return an error from Harmony. Callers are encouraged to check get_transformation_job_status first if the job’s current state is uncertain.

Input Parameters

Status	MCP Argument	harmony-py Parameter	Description
✅	`job_id`	`job_id`	The ID of the Harmony job to control.
✅	`action`	N/A (dispatches to `cancel()` / `pause()` / `resume()`)	One of `"cancel"`, `"pause"`, or `"resume"`.

Output Fields

Same output model as get_transformation_job_status (GetTransformationJobStatusOutput) — see that section’s Output Fields table for the full field list (job_id, status, message, progress, created_at/updated_at, links, total_hits, next_cursor, code, description, etc.). The returned status reflects the job’s state immediately after the requested action is applied (e.g., canceled, paused, running), not a final/terminal outcome if the action triggers further async processing. code/description are populated if input validation fails, if the control action itself fails (e.g., invalid state transition), or if Harmony’s status response doesn’t match the expected schema.