MCP App Store
by KeenEthics
Rescale icon

Rescale MCP Server for Claude and ChatGPT

By Rescale

Overview

Connect Claude to your Rescale workspace to inspect and manage high-performance computing resources using natural language. Find and monitor jobs, review logs and results, discover available software and hardware, manage files and workflows, and work with supported storage, budget, and data-connector features. The connector uses Rescale OAuth and respects the authenticated user's existing organization, workspace, and resource permissions. Available capabilities depend on the user's Rescale plan, region, workspace access, and enabled products.

How to connect

  1. Claude: Open the Rescale listing in the Claude connectors directory and choose Connect. Claude listing Rescale exposes 114 tools here.
  2. ChatGPT: Open the Rescale listing in the ChatGPT app directory and choose Connect. ChatGPT listing Rescale exposes 114 tools here.

Tools

add_file_tag

ChatGPT
Add a new or existing tag to a file.

add_job_tag

ChatGPT
Add a new tag to a job.

add_skills_to_profile

ChatGPT
An agent profile is a Rescale Agentic Project. Attach one or more agent skills to an Agentic Project. Only the project's owner may do this, and every skill id must belong to a skill in your workspace. Returns the updated project with its nested skills. Path Parameters: - id: Obfuscated Agentic Project id

analyze_deposit_consumption

ChatGPT
Analyze deposit consumption with optional breakdowns by user, workspace, or charge type. This tool provides detailed consumption analysis for a specific deposit, allowing you to understand how the deposit balance has been used over time and by different dimensions. Args: deposit_id: The customer-facing (obfuscated) deposit ID (e.g., "ABCDE", "nkaaa") This is the ID shown to customers, not the internal database ID breakdown_by: Optional breakdown dimension: - "user": Group consumption by user/member - "workspace": Group consumption by workspace - "charge_type": Group consumption by charge type (hardware, software, storage, etc.) - None: Return total consumption without breakdown start_month: Start month in YYYY-MM format (e.g., "2025-08") If not provided, returns all available history end_month: End month in YYYY-MM format (e.g., "2025-12") If not provided, returns up to current month org_code: Organization code (auto-detects from user context if not provided) Returns: Dictionary containing: - deposit_id: The deposit ID analyzed - consumption: List of consumption records with breakdown if requested - breakdown_by: The breakdown dimension used (or None) - start_month: Start of date range - end_month: End of date range - total_consumed: Total amount consumed in the period - org_code: Organization code Example usage: Total consumption without breakdown analyze_deposit_consumption(deposit_id="ABCDE") Consumption breakdown by user for last month analyze_deposit_consumption( deposit_id="ABCDE", breakdown_by="user", start_month="2026-01", end_month="2026-01" ) Consumption breakdown by charge type for August 2025 analyze_deposit_consumption( deposit_id="ABCDE", breakdown_by="charge_type", start_month="2025-08", end_month="2025-08" ) Consumption trend over 3 months analyze_deposit_consumption( deposit_id="ABCDE", start_month="2025-10", end_month="2025-12" )

ask_rescale_agent

ChatGPT
Ask Rescale Copilot a natural-language question. Copilot orchestrates Rescale's specialist agents (jobs, billing, data, and registered A2A agents) and returns a synthesized answer. Use for open-ended questions about your Rescale data, jobs, or account that a single API call can't answer. The question and answer are saved as a conversation in the user's Rescale Assistant history. Without conversation_id, each call starts a new conversation; pass the conversation_id from a previous answer to continue that thread. Args: question: The natural-language question to ask. conversation_id: Optional prior conversation id to continue a thread. Returns: Dict with success, markdown (the answer), conversation_id, and citations. Inline components appear as text summaries. Interactions that require a client response return success=False.

assign_job_custom_fields

ChatGPT
This will create custom fields specific to a job. For creating custom fileds that can hold multiple values use isMultiple flag while creating custom fields. Use submit_job_custom_fields to set values on these fields.

attach_job_to_storage_device

ChatGPT
Attach a job's output files to the storage device Path Parameters: - id: Storage device ID

check_file_dependencies

ChatGPT
Check if LS-DYNA and Abaqus simulation files contain all required include files. Downloads and analyzes any combination of archives (.zip/.tar/.7z) and standalone deck files (.dyn/.k/.key/.inp) to find INCLUDE statements and validate that all referenced files are present across the combined file set. Args: file_ids: List of Rescale file IDs to download and validate together. Can be a single archive or a mix of archives and standalone files. Returns: Dict containing: - success: Boolean indicating if analysis completed - total_files: Total number of files across all sources - deck_files_found: Number of simulation input files found - deck_files: List of simulation input files (.k/.key/.dyn/.inp) - required_filecount: Number of files referenced by INCLUDE - required_files: List of files referenced by *INCLUDE statements - missing_files: List of missing files (null if complete) - is_complete: Boolean indicating if all dependencies present - found_files: Number of dependencies successfully found

create_agent_profile

ChatGPT
Create an Agentic Project (called an agent profile in the API) owned by the authenticated user. Attach skills afterward with add_skills_to_profile. Requires the Agentic Engineering AI tier or higher.

create_agent_skill

ChatGPT
Create an agent skill owned by the authenticated user. Requires the Agentic Engineering AI tier or higher.

create_budget

ChatGPT
Create a new budget. Budget levels determine scope and target format: Organization (array of workspace IDs like ['00-046880176', '00-049568838']), Workspace (single workspace ID string like '00-046880176'), Project (project ID string like 'pCTMk'), Member (user ID string like 'pCTMk'). Budget units: 'cost' for currency amounts (e.g., '$100.00', '€75.50'), 'core-hour' for compute time (e.g., '1000'). Budget types: 'One-time' or 'Periodic' (Monthly/Quarterly/Annual). Charge types: 'All', 'Hardware', 'Software', 'Compute & License'.

create_data_connector_reference

ChatGPT
Create a workspace-scoped handle to a Data Connector Service (DCS) connector so it can be linked to an Agentic Project. Get the connector_id from dcs_list_connectors first (only reference connectors whose status is 'active'). Idempotent on (connector_id, connector_path): re-creating an existing handle updates its cached display fields and returns 200 instead of 201. Returns the reference, whose obfuscated id is passed as entity_id to link_entity_to_agent_profile with entity_type 'aiphysics.dataconnectorreference'. Requires the Agentic Engineering AI tier or higher.

create_job

ChatGPT
Create a new saved job with its software, hardware, and input files. The job does not start running until it is submitted with submit_job. If the hardware type is omitted, it is taken from the selected software; interactive can be set for workstations.

create_job_comment

ChatGPT
Add a comment to a job, or a reply to an existing comment when parentId is set. Posting emails the job owner and earlier commenters on the job who have comment notifications on, but not the commenter. Every address written as @email in the comment (for example @name@example.com) is emailed a mention notice, unless it belongs to a Rescale user who turned those notices off.

create_storage_device

ChatGPT
Create a new storage device with hardware configuration. Hardware configuration and storageSize can be obtained from get_storage_device_pricing tool.

create_workflow

ChatGPT
Create a new workflow template from a complete caller-provided WorkflowMarkup definition. This tool does not translate a natural-language request into WorkflowMarkup. The platform validates the definition and returns structured validation errors for invalid input.

dcs_browse

ChatGPT
List the files and folders at a path inside a Data Connector Service (DCS) connector's data source. Use this to find the connector_path to scope a data connector reference to a single file or folder. Folders (type 'folder', has_children true) can be browsed into by passing their path back as the path parameter.

dcs_download_file

ChatGPT
Download a complete file from a Data Connector Service (DCS) connector as base64. Use the connectorId and a FILE path from get_data_connector_reference, or a file path returned by dcs_browse. Folders cannot be downloaded; browse them and download each file. Files larger than the inline download limit are rejected; use dcs_preview_file to read part of them. Returns success, connector_id, path, filename, binary_content (base64), and size.

dcs_get_adx_schema_profile

ChatGPT
Get the column schema and a sampled data profile for one table of an Azure Data Explorer (ADX) connector in the Data Connector Service (DCS). Call this before writing a Kusto Query Language (KQL) query or planning analysis. A partial status with issues means some observations are missing; sampled population and datetime statistics may have more source rows (check source_has_more).

dcs_list_connectors

ChatGPT
List the Data Connector Service (DCS) connectors in the user's workspace. A connector links Rescale to one of the workspace's external data sources: Amazon S3, Azure Blob Storage, Google Cloud Storage, Databricks, SharePoint, Jama, or Azure Data Explorer (ADX). Results are paginated with limit and offset. The list is empty when the workspace has no connectors; the other dcs_ tools need a connector or documents indexed from one.

dcs_list_documents

ChatGPT
List or count the documents that the Data Connector Service (DCS) has indexed from the workspace's connectors, using the metadata captured when each file was ingested. Use this when the question is about the documents themselves rather than their contents: which documents exist, how many there are, which was modified or created most recently, which are largest, or which changed inside a date range. It covers every indexed document, so a superlative such as 'the most recently modified document' applies to all of them. When list_info.version_families.collapsed is greater than 0, revisions were collapsed within the returned page by filename version or date, not by the requested order, so a collapsed revision may have been the largest or newest. dcs_search_semantic can only order the documents its query retrieved, so the same question answered there is true only of that sample. This tool takes no query and cannot search document contents; questions about what a document says are answered by dcs_search_semantic.

dcs_preview_file

ChatGPT
Read the first bytes of a file from a Data Connector Service (DCS) connector. Use the connectorId and connectorPath of a FILE linked entity retrieved with get_data_connector_reference (pass connectorPath as path). A FOLDER or whole-connector link must be browsed with dcs_browse to find a file first. This is a preview, not the full file: it reads 32 KiB by default and at most 1 MiB; use dcs_download_file for the complete file. Use format=json to receive content and preview metadata together. Only text files can be reliably read as text; binary formats are not converted to readable text.

dcs_query_adx

ChatGPT
Run one read-only Kusto Query Language (KQL) query against the database of an Azure Data Explorer (ADX) connector in the Data Connector Service (DCS). Get id from dcs_list_connectors and table and column names from dcs_get_adx_schema_profile. DCS limits KQL to 20,000 UTF-8 bytes. Results are never truncated: exceeding 1,000 rows, 256 columns, or 3 MiB of data fails (502, ADX_QUERY_LIMIT_EXCEEDED), and responses over 4 MiB fail (502, mcp_response_too_large). Bound queries with take, summarize, or project. DCS times out at 60 s, and the whole call at 75 s (504). Management commands, set statements, non-let intermediate statements, multiple final query expressions, cluster()/database() references, externaldata, unsupported plugins, and callouts are rejected (400, ADX_QUERY_POLICY_REJECTED). Syntax and semantic errors return 422. Handled query failures are JSON with status, error, details, and when present code, reason, retryable, retry_after, hint, request_id. On 429 wait retry_after seconds if present; mcp_capacity_timeout means this server was at capacity and DCS was not contacted.

dcs_search_agentic

ChatGPT
Search document contents across the workspace's Data Connector Service (DCS) connectors with multi-step retrieval: the query is broken into sub-queries, and ranked evidence passages are returned. When include_citations is true, the response also includes structured source data for attribution. Citation sources report a score of 0 because results are ranked by position, not scored. Available only when multi-step retrieval is enabled in the DCS deployment.

dcs_search_semantic

ChatGPT
Search document contents across the workspace's Data Connector Service (DCS) connectors with a natural-language query. Results are limited to the user's workspace. Use this when the question is about what a document says. For questions about the documents themselves (which exist, how many, newest, largest, changed in a date range) use dcs_list_documents: this tool can only order the documents its query retrieved, so a superlative it produces is true of that sample rather than of all documents. When include_citations is true, returns structured citation sources with stable IDs for attribution.

decide_linked_entity

ChatGPT
Record the project owner's ACCEPT or REJECT decisions on links that are PROPOSED. Call only when the owner has clearly decided on specific proposals (for example: 'yes, link that job', 'no to the file'), naming them in this conversation. Never call on your own initiative. Send every decision the owner made in one call. Returns one result per decision, in request order; check each result's ok, since one decision failing does not stop the others. A proposed link only takes effect after ACCEPT. REJECT is not delete: the object is never proposed to this project again, but the owner can still link it by asking. Accepting a link that is already ACTIVE succeeds and changes nothing. Only the project owner may decide. Path Parameters: - id: Injected server-side by the caller. Do not set.

delete_agent_profile

ChatGPT
An agent profile is a Rescale Agentic Project. Soft-delete an Agentic Project (deactivates it; the project stops appearing in every Agentic Project endpoint). Only the project's owner may delete it. Deactivation is not reversible via the API. Path Parameters: - id: Obfuscated Agentic Project id

delete_agent_skill

ChatGPT
Delete an agent skill. Only the skill's owner may delete it. Path Parameters: - id: Obfuscated agent skill id

delete_budget

ChatGPT
Delete an existing budget. Requires the budget ID in the request body. This operation permanently removes the budget and cannot be undone. Users can only delete budgets they have permission to manage based on their organizational role.

delete_file

ChatGPT
Delete a file from storage

delete_job

ChatGPT
Delete a saved job (cannot delete running jobs)

delete_storage_device_file

ChatGPT
Delete a file at the specified path on the storage device Path Parameters: - id: Storage device ID - path: File path to delete

delete_workflow

ChatGPT
Soft-delete (archive) a workflow template; its workflow runs are kept. Fails while any of its workflow runs is queued or running — stop them first via stop_workflow_run. Path Parameters: - workflowId: Obfuscated workflow template ID

delete_workflow_run

ChatGPT
Soft-delete (archive) a workflow run. Fails with 400 if the run is still running — call stop_workflow_run first and wait for it to reach a terminal status before deleting. Path Parameters: - workflowRunId: Obfuscated workflow run ID

download_file

ChatGPT
Download file content. Supports HTTP Range requests (RFC7233) for partial downloads. Files larger than 5GB should use Rescale CLI or Transfer Manager.

generate_billing_csv

ChatGPT
For organization and workspace admins; other users get a permission error. Generate billing CSV and return a presigned S3 download URL. Organization is derived from auth — do not pass it. Max 12-month date range. Required params by type: 'organization' (no extra params), 'workspace' (workspace if org admin), 'member' (member email + workspace if org admin), 'project' (project ID + workspace if org admin). Workspace admins never need to pass workspace — it is auto-detected.

get_agent_profile

ChatGPT
Retrieve a single Agentic Project (called an agent profile in the API), including its nested skills, by its obfuscated id. Visible only to the owner or, when shared, users in the same workspace. Path Parameters: - id: Obfuscated Agentic Project id

get_agent_skill

ChatGPT
Retrieve a single agent skill by its obfuscated id. Any user in the owner's company may read it. Path Parameters: - id: Obfuscated agent skill id

get_analysis_details

ChatGPT
Retrieves details about an analysis including license configuration settings. This endpoint includes license settings as part of the analysis response when available. Path Parameters: - code: Unique code identifier for the analysis

get_billing_summary

ChatGPT
Get an organization's or workspace's billing summary, for organization and workspace admins. Returns the same multicurrency spending totals shown on the admin billing page, optionally broken down by workspace, project, or member and limited to a month range. Both context_id and context_type are required; call get_user_context first to find the values available to the user. Args: context_id: REQUIRED - Organization code (e.g. 'acme') or 'org_code/workspace_id' format context_type: REQUIRED - "workspace" or "organization" aggregate_by: "workspaces" (organization context only), "projects", "members", or None for basic totals start_year: Start year for date filtering (e.g. 2025) start_month: Start month for date filtering (e.g. 12) end_year: End year for date filtering (e.g. 2025) end_month: End month for date filtering (e.g. 12) currency: Currency code for financial data (default: USD) Returns: Dictionary containing billing summary data matching the UI format: - Organization billing: Aggregated multicurrency billing summary - Workspace billing: Individual workspace multicurrency billing - Includes aggregate_by and context (the context_id that was summarized) Example usage: get_billing_summary("acme", "organization") Organization billing for 'acme' get_billing_summary("acme/00-046880176", "workspace") Specific workspace get_billing_summary("acme", "organization", start_year=2025, start_month=12, end_year=2025, end_month=12) get_billing_summary("acme", "organization", currency="EUR") Organization billing in EUR

get_data_connector_reference

ChatGPT
Get the connector UUID and file or folder path for a data connector reference. Pass the entity_id from a list_agent_profile_linked_entities result whose entity_type is aiphysics.dataconnectorreference. For a FILE reference, pass connectorId and connectorPath (as path) to dcs_download_file (or dcs_preview_file for a quick look). For a FOLDER or whole-connector reference, browse connectorPath (or /) with dcs_browse, including subfolders and every page, to find its files. Requires access to the reference's workspace and the Agentic Engineering AI tier. Path Parameters: - id: Obfuscated data connector reference entity_id from an Agentic Project link

get_file

ChatGPT
Get metadata for a specific file

get_file_download_status

ChatGPT
Get the status of a file download operation Path Parameters: - id: Storage device ID - fdId: File download ID

get_file_lines

ChatGPT
Retrieve the content of a plain text file as an array of lines. Files larger than 5GB should use Rescale CLI or Transfer Manager.

get_folders

ChatGPT
Retrieve the complete folder hierarchy for the authenticated user, including home folders (personal), workspace folders (shared with team), and folders shared with the user. Returns a nested tree structure with all subfolders at every level, allowing you to navigate the folder organization, find folders by name, build folder paths, and understand the user's folder permissions. Use this endpoint to discover available folders.

get_job

ChatGPT
Get detailed information about a specific job

get_job_comments

ChatGPT
Retrieves all comments and nested replies for a specific job. Comments are returned in a hierarchical structure with children nested recursively. Path Parameters: - jobId: Obfuscated job ID

get_job_data

ChatGPT
Filter and aggregate many jobs for analytics and reporting. Requires Rescale Data Intelligence to be enabled for the user's organization; otherwise it returns an error saying it is not enabled. Returns job metadata including execution details, software, hardware, status, and custom fields, without billing data. Jobs can be filtered by software, status, hardware values, tags, folders, and custom field values. Results are ordered newest first and cannot be sorted. Listing recent jobs or looking up one job is done with list_jobs or get_job, which need no extra entitlement.

get_job_download_status

ChatGPT
Get the status of a job download operation Path Parameters: - id: Storage device ID - jobDownloadId: Job download ID

get_job_extended_logs

ChatGPT
Get job logs including statuses and user logs.

get_job_files

ChatGPT
Get all output files for a completed job

get_job_input_files

ChatGPT
Get input files attached to a Rescale job. Fetches the job detail and extracts input file metadata from each job analysis. Returns compact file info (id, name, size) for every unique input file. A job's output files are returned by get_job_files. Args: job_id: The Rescale job ID Returns: Dict with success, job_id, count, and files list

get_job_metrics

ChatGPT
Get CPU, memory, and other performance metrics for a job. Only returns data if the job is currently in executing state.

get_job_run

ChatGPT
Get details for a specific job run

get_job_runs

ChatGPT
Get all runs for a job (DOE jobs have multiple runs)

get_job_statuses

ChatGPT
Get the status history for a job

get_run_files

ChatGPT
Get output files for a specific run

get_storage_device

ChatGPT
Get details of a specific storage device Path Parameters: - id: Storage device ID

get_storage_device_logs

ChatGPT
Get user logs for a storage device Path Parameters: - id: Storage device ID

get_storage_device_pricing

ChatGPT
Calculate pricing for storage device configuration

get_storage_device_shared_users

ChatGPT
Get list of users the storage device is shared with Path Parameters: - id: Storage device ID

get_storage_device_statuses

ChatGPT
Get status history for a storage device Path Parameters: - id: Storage device ID

get_user_context

ChatGPT
Get current user's context including organization and workspace information. Retrieves user profile from /api/v1/users/me/ to determine: - Current organization code and details - Current workspace ID and details - User's admin permissions (org admin, workspace admin) - Suggested context for other API calls Returns: Dictionary containing: - user_info: Basic user information (user_id, email, full_name) - organization: Organization details (code, name, display_name) - workspace: Current workspace details (external_id, name) - permissions: Admin status flags - suggested_contexts: context_id values the user can pass to billing and other admin tools - default_context: The first entry of suggested_contexts, or null if there are none Example usage: user_context = get_user_context() org_code = user_context["organization"]["code"] is_admin = user_context["permissions"]["is_organization_admin"]

get_user_info

ChatGPT
Retrieves user profile information such as ID, email, ssh ports, organization code, and more.

get_user_license_info

ChatGPT
Get the current user's license options for one simulation software package: their saved license settings, the on-demand licenses they can use, and the license providers available for that software. Use the result to set up licensing for a job. This only reads license information and does not buy or change any license.

get_workflow

ChatGPT
Retrieve a workflow template with its latest definition, version number, and metadata. Path Parameters: - workflowId: Obfuscated workflow template ID (not a workflow run ID)

get_workflow_run

ChatGPT
Retrieve a workflow run by id: its status history, its workflow template's workflowId, and the graph it ran with, including each node's execution status. A Completed run may have failed or been stopped; statusReason says which. Each node's resourceType and resourceId identify the job or storage device it ran. Path Parameters: - workflowRunId: Obfuscated workflow run ID

list_agent_profile_activity

ChatGPT
An agent profile is a Rescale Agentic Project. List the project's activity log, newest-first and paginated. Readable by anyone who can see the project. Path Parameters: - id: Obfuscated Agentic Project id

list_agent_profile_linked_entities

ChatGPT
An agent profile is a Rescale Agentic Project. List the jobs, files, and workflows linked to the project, newest-first and paginated. Each result carries non-sensitive metadata (e.g. name, hardware) about the linked object without granting access to it. Readable by anyone who can see the project. Path Parameters: - id: Obfuscated Agentic Project id

list_agent_profile_usage

ChatGPT
An agent profile is a Rescale Agentic Project. List the project's usage log (who used it and when), newest-first and paginated. Readable by anyone who can see the project. Path Parameters: - id: Obfuscated Agentic Project id

list_agent_profiles

ChatGPT
List Agentic Projects (called agent profiles in the API) visible to you: those you own plus any shared with your workspace. Requires the Agentic Engineering AI tier or higher; without it every Agentic Project endpoint returns 404. Each project includes its nested skills. Results are paginated and ordered newest-first.

list_agent_skills

ChatGPT
List agent skills owned by users in your company. Requires the Agentic Engineering AI tier or higher; without it every agent skill endpoint returns 404. Results are paginated and ordered newest-first.

list_automations

ChatGPT
Get a list of available automations with details including environment variables, dependencies, and command.

list_budgets

ChatGPT
Get all budgets accessible to the authenticated user. Returns different budgets based on user permissions: regular users see only their own budgets, workspace admins see workspace/member/project budgets, organization admins see all budgets in the organization including multi-workspace budgets. Budget levels include: Organization (spans multiple workspaces), Workspace (single workspace), Project (specific project), Member (specific user).

list_core_types

ChatGPT
Get all available hardware core types with pricing and specifications. Set prefetch=true and fepf=true.

list_credits

ChatGPT
For organization and workspace admins; other users get a permission error. List credits for organization or workspace with automatic context detection and optional expiration filtering. Consolidates organization and workspace credit queries into a single tool. Auto-detects context or uses explicit parameters to determine scope. Args: context_id: Organization code (e.g., 'acme') or workspace ID (e.g., '00-046880176') If not provided, auto-detects from user context context_type: "organization" or "workspace" - explicitly set the context level If not provided, auto-detects based on user permissions types: Comma-separated credit types: software, hardware, csp-credit-transfer, platform-credit Defaults to all types if not specified providers: Comma-separated cloud providers: amazon, google, microsoft expiring_within_days: If provided, only return credits expiring within this many days Useful for finding credits that need attention soon Returns: Dictionary containing: - credits: List of credits with details (id, amount, remaining, type, expiration, etc.) - count: Total number of credits (after filtering if expiring_within_days provided) - context_type: Whether organization or workspace level - context_id: The organization/workspace queried - filtered_by_expiration: Boolean indicating if expiration filter was applied - days_until_expiration: Added to each credit when filtering by expiration Example usage: list_credits() Auto-detect user's context, all credits list_credits(context_type="organization") Organization credits list_credits(context_id="acme", context_type="organization") Specific org list_credits(context_id="00-046880176", context_type="workspace") Specific workspace list_credits(types="software,hardware") Filter by types list_credits(expiring_within_days=30) Only credits expiring in next 30 days list_credits(expiring_within_days=7) Credits expiring within a week

list_deposits

ChatGPT
List deposits for organization or workspace with automatic context detection and optional expiration filtering. Consolidates organization and workspace deposit queries into a single tool. Auto-detects context or uses explicit parameters to determine scope. Note: Requires partner/admin permissions to view deposits. Args: context_id: Organization code (e.g., 'acme') or workspace ID (e.g., '00-046880176') If not provided, auto-detects from user context context_type: "organization" or "workspace" - explicitly set the context level If not provided, auto-detects based on user permissions expiring_within_days: If provided, only return deposits expiring within this many days Useful for finding deposits that need attention soon Returns: Dictionary containing: - deposits: List of deposits with details (id, amount, remaining, effective_date, expiration, etc.) - count: Total number of deposits (after filtering if expiring_within_days provided) - context_type: Whether organization or workspace level - context_id: The organization/workspace queried - filtered_by_expiration: Boolean indicating if expiration filter was applied - days_until_expiration: Added to each deposit when filtering by expiration Example usage: list_deposits() Auto-detect user's context, all deposits list_deposits(context_type="organization") Organization deposits list_deposits(context_id="acme", context_type="organization") Specific org list_deposits(context_id="00-046880176", context_type="workspace") Specific workspace

list_file_tags

ChatGPT
List all tags on a file.

list_files

ChatGPT
Get all files owned by the user. Set owner=1 to exclude files shared with the user.

list_folder_contents

ChatGPT
List the files and subfolders directly inside a Rescale Files folder: subfolders first, then files, each newest first. Each result has a type ('file' or 'folder') and an item; pass a file item's id to get_file_lines or download_file, and a folder item's id back to this tool to go one level deeper. Use it for a linked folder on an Agentic Project (entity type 'jobs.filefolder'). Results are cursor-paginated: when 'next' is set, pass its cursor query value as cursor to get the next page. Path Parameters: - folder_id: Obfuscated id of the folder

list_job_custom_fields

ChatGPT
Get custom field definitions and values for a job.

list_job_instances

ChatGPT
Get VM instances for a job.

list_job_tags

ChatGPT
List all tags on a job.

list_jobs

ChatGPT
List jobs owned by or shared with the authenticated user, with basic filtering by job type, status, and text search. Use this to list recent jobs or find a job by name; get_job returns one job's details.

list_minimal_analyses

ChatGPT
List the software packages (called analyses in the Rescale API) available to run on Rescale, with a few fields per package. Pass a package's code to get_analysis_details for its full details and license settings. fields picks the returned fields, for example 'code,name'.

list_organization_workspaces

ChatGPT
Get all workspaces (accounts) within an organization. This endpoint lists all workspaces that the authenticated user has access to within the specified organization. Organization admins can see all workspaces, while workspace admins can only see workspaces they administer. Essential for getting workspace spending breakdowns. Path Parameters: - code: Organization code (e.g., 'rescale', 'acme')

list_run_directory_contents

ChatGPT
Get files from a run's working directory (in-progress runs only)

list_storage_device_files

ChatGPT
List files and directories at the specified path on the storage device Path Parameters: - id: Storage device ID - path: File path on the storage device (defaults to root / if not specified)

list_storage_device_regions

ChatGPT
Get list of available regions for storage devices

list_storage_devices

ChatGPT
Get a paginated list of storage devices

list_workflow_runs

ChatGPT
List workflow runs, the executions of workflow templates, with filtering and pagination.

list_workflows

ChatGPT
List workflow templates owned by or shared with the authenticated user, plus templates published to all users.

remove_file_tag

ChatGPT
Remove a tag from a file.

remove_job_tag

ChatGPT
Remove a specific tag from a job.

remove_skills_from_profile

ChatGPT
An agent profile is a Rescale Agentic Project. Detach one or more agent skills from an Agentic Project. Only the project's owner may do this, and every skill id must belong to a skill in your workspace. Returns the updated project with its remaining skills. Path Parameters: - id: Obfuscated Agentic Project id

resume_job

ChatGPT
Resume a suspended interactive workstation job. Its cloud hardware starts again, and compute time is billed to the user's Rescale account while it runs.

share_agent_profile_with_workspace

ChatGPT
An agent profile is a Rescale Agentic Project. Make the Agentic Project readable by everyone in your workspace. Only the project's owner may share it. Returns the updated project. Path Parameters: - id: Obfuscated Agentic Project id

stop_job

ChatGPT
Gracefully stop a running job and save output files

stop_storage_device

ChatGPT
Shutdown/stop a running storage device Path Parameters: - id: Storage device ID

stop_workflow_run

ChatGPT
Stop a workflow run that is in queued or running state. Stops all running nodes (jobs, workstations, storage devices). The run shows Stopping until it reaches Completed. Path Parameters: - workflowRunId: Obfuscated workflow run ID

submit_job

ChatGPT
Submit a saved job to start it. Submitting starts the job's cloud hardware, and its compute time is billed to the user's Rescale account until the job finishes or is stopped.

submit_job_custom_fields

ChatGPT
Submit custom field values for a job. This is a custom tool that works around FastMCP's limitation with additionalProperties in OpenAPI schemas. Args: job_id: The job ID to update custom fields for custom_fields: Dictionary mapping field names to values. Values can be strings, numbers, booleans, or string arrays. Example: { "test_context_field": "Some text value", "test_input_field": 777, "temperature": 25.5, "converged": true, "tags": ["simulation", "validation"] } Returns: Dictionary with success status and API response

submit_storage_device

ChatGPT
Submit a storage device to provision it. Provisioning starts the device's cloud hardware and storage, which are billed to the user's Rescale account while the device runs. get_storage_device_pricing shows the rates.

submit_workflow

ChatGPT
Start a workflow run from the workflow template's latest version. Parameters are validated against the template's parameter definitions. Job steps in the workflow submit Rescale jobs, and their compute time is billed to the user's Rescale account. Limited to 250 submissions per hour. Returns the new workflow run.

suspend_job

ChatGPT
Suspend a running job to pause execution and save current state

sync_files_to_storage_device

ChatGPT
Initiate file download/sync to the storage device Path Parameters: - id: Storage device ID

unshare_agent_profile_from_workspace

ChatGPT
An agent profile is a Rescale Agentic Project. Revert the Agentic Project to private (owner-only). Only the project's owner may unshare it. Returns the updated project. Path Parameters: - id: Obfuscated Agentic Project id

update_agent_profile

ChatGPT
Partially update an Agentic Project (called an agent profile in the API). Only the project's owner may modify it. To change the attached skills, use add_skills_to_profile or remove_skills_from_profile.

update_agent_skill

ChatGPT
Partially update an agent skill. Only the skill's owner may modify it. Path Parameters: - id: Obfuscated agent skill id

update_budget

ChatGPT
Update an existing budget. The budget ID must be included in the request body along with the fields to update. Updatable fields include: budgeted_amount (currency format for cost budgets, numeric for core-hour budgets), name (max 50 characters), charge_type ('All', 'Hardware', 'Software', 'Compute & License'), budget_type ('One-time', 'Periodic'), budget_period ('Monthly', 'Quarterly', 'Annual' - required for Periodic budgets), analysis (required when charge_type is 'Software'), start_date/end_date (date format), strict_budget (boolean).

update_job

ChatGPT
Change the configuration of a saved job, such as its name, software, hardware, or input files. The provided fields overwrite the job's current values. After a job is submitted, its configuration cannot be changed; only details such as its name, description, and priority can be updated.

update_storage_device

ChatGPT
Update an existing storage device Path Parameters: - id: Storage device ID

update_workflow

ChatGPT
Update a workflow template's name, description, and/or definition. Providing the workflow field creates a new WorkflowVersion with an incremented version number; the field is an opaque WorkflowMarkup blob validated by the platform on submit. All fields are optional. Path Parameters: - workflowId: Obfuscated workflow template ID

upload_file

ChatGPT
Upload a file to the user's Rescale files. Each call creates a new file; an existing file with the same name is not replaced. file_content is the content as plain text, or base64 for binary data: content that is valid base64 is stored as the decoded bytes, anything else as UTF-8 text. filename is the file's name; executable and script types such as .exe, .bat, .js, and .jar are rejected. folder_id is an optional folder ID from get_folders for the file's folder. file_type_id is accepted for compatibility and has no effect. Uploads time out after 60 seconds.

add_file_tag

Claude

add_job_tag

Claude

add_skills_to_profile

Claude

agent_runtime_download

Claude

analyze_deposit_consumption

Claude

assign_job_custom_fields

Claude

attach_job_to_storage_device

Claude

check_ai_quota

Claude

check_file_dependencies

Claude

create_agent_profile

Claude

create_agent_skill

Claude

create_budget

Claude

create_data_connector_reference

Claude

create_job

Claude

create_job_comment

Claude

create_storage_device

Claude

create_workflow

Claude

dcs_browse

Claude

dcs_list_connectors

Claude

dcs_list_documents

Claude

dcs_search_agentic

Claude

dcs_search_semantic

Claude

delete_agent_profile

Claude

delete_agent_skill

Claude

delete_budget

Claude

delete_file

Claude

delete_job

Claude

delete_storage_device_file

Claude

delete_workflow

Claude

delete_workflow_run

Claude

download_binary_file

Claude

download_file

Claude

fetch_pptx_template

Claude

generate_billing_csv

Claude

get_agent_profile

Claude

get_agent_skill

Claude

get_analysis_details

Claude

get_billing_summary

Claude

get_file

Claude

get_file_download_status

Claude

get_file_lines

Claude

get_folders

Claude

get_job

Claude

get_job_comments

Claude

get_job_data

Claude

get_job_download_status

Claude

get_job_extended_logs

Claude

get_job_files

Claude

get_job_metrics

Claude

get_job_run

Claude

get_job_runs

Claude

get_job_statuses

Claude

get_run_files

Claude

get_storage_device

Claude

get_storage_device_logs

Claude

get_storage_device_pricing

Claude

get_storage_device_shared_users

Claude

get_storage_device_statuses

Claude

get_user_context

Claude

get_user_info

Claude

get_user_license_info

Claude

get_workflow

Claude

get_workflow_run

Claude

list_agent_profile_activity

Claude

list_agent_profile_linked_entities

Claude

list_agent_profile_usage

Claude

list_agent_profiles

Claude

list_agent_skills

Claude

list_automations

Claude

list_budgets

Claude

list_core_types

Claude

list_credits

Claude

list_deposits

Claude

list_file_tags

Claude

list_files

Claude

list_job_custom_fields

Claude

list_job_instances

Claude

list_job_tags

Claude

list_jobs

Claude

list_minimal_analyses

Claude

list_organization_workspaces

Claude

list_run_directory_contents

Claude

list_storage_device_files

Claude

list_storage_device_regions

Claude

list_storage_devices

Claude

list_tool_capabilities

Claude

list_workflow_runs

Claude

list_workflows

Claude

record_agent_profile_activity

Claude

record_agent_profile_usage

Claude

remove_file_tag

Claude

remove_job_tag

Claude

remove_skills_from_profile

Claude

resume_job

Claude

share_agent_profile_with_workspace

Claude

stop_job

Claude

stop_storage_device

Claude

stop_workflow_run

Claude

submit_job

Claude

submit_job_custom_fields

Claude

submit_storage_device

Claude

submit_workflow

Claude

suspend_job

Claude

sync_files_to_storage_device

Claude

unshare_agent_profile_from_workspace

Claude

update_agent_profile

Claude

update_agent_skill

Claude

update_budget

Claude

update_job

Claude

update_storage_device

Claude

update_workflow

Claude

upload_file

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/mcp-services-platform-rescale-com"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/mcp-services-platform-rescale-com)