generated: '2026-07-19' method: derived status: candidate source: https://github.com/lifebit-ai/cloudos-cli notes: Lifebit publishes no hosted or remote MCP server. Searches of the lifebit-ai GitHub organisation, the npm @modelcontextprotocol namespace and the public MCP registries returned no first-party server. This is a CANDIDATE tool list only — an API Evangelist proposal, not a Lifebit claim. Because Lifebit publishes no OpenAPI description, the candidate tools are grounded in the REST endpoints exercised by the first-party open-source client, cloudos-cli, and each tool names the endpoint it would wrap. Lifebit does market an "agentic federated TRE" product, but exposes no public agent protocol endpoint for it. server: name: lifebit-cloudos transport: http url: null base_api: https://cloudos.lifebit.ai/api auth: type: apiKey in: header name: apikey tools: - name: get_current_user description: Return the authenticated user and their workspaces. source_endpoint: GET /api/v1/users/me - name: list_projects description: List workspace projects, with paging and name search. source_endpoint: GET /api/v2/projects?teamId={workspaceId} - name: create_project description: Create a new project in the workspace. source_endpoint: POST /api/v1/projects - name: list_workflows description: List available pipelines/workflows in the workspace. source_endpoint: GET /api/v3/workflows?teamId={workspaceId} - name: list_job_queues description: List job queues available for a workflow type. source_endpoint: GET /api/v1/jobs?teamId={workspaceId} - name: submit_job description: Submit a Nextflow job to a project and queue. source_endpoint: POST /api/v2/jobs - name: get_job description: Fetch a single job, including status and metadata. source_endpoint: GET /api/v1/jobs/{jobId}?teamId={workspaceId} - name: list_jobs description: List jobs in the workspace, filterable by status and archive state. source_endpoint: GET /api/v2/jobs?teamId={workspaceId} - name: get_job_results description: Fetch result data for a completed job. source_endpoint: GET /api/v1/jobs/{jobId}/data - name: abort_job description: Abort a running or initialising job. source_endpoint: PUT /api/v2/jobs/{jobId}/abort - name: list_interactive_sessions description: List interactive analysis sessions. source_endpoint: GET /api/v3/interactive-sessions - name: create_interactive_session description: Provision an interactive analysis session. source_endpoint: POST /api/v1/interactive-sessions - name: resume_interactive_session description: Resume a paused interactive session. source_endpoint: PUT /api/v1/interactive-sessions/{sessionId}/resume - name: abort_interactive_session description: Stop an interactive session. source_endpoint: PUT /api/v1/interactive-sessions/{sessionId}/abort - name: list_datasets description: List datasets in the workspace File Explorer. source_endpoint: GET /api/v2/datasets - name: list_dataset_items description: List files and folders inside a dataset folder. source_endpoint: GET /api/v1/datasets/{folderId}/items - name: move_data_items description: Move files or folders between locations. source_endpoint: PUT /api/v1/dataItems/move - name: list_s3_bucket_contents description: Browse a linked S3 bucket path. source_endpoint: GET /api/v1/data-access/s3/bucket-contents - name: list_azure_container_contents description: Browse a linked Azure Blob container path. source_endpoint: GET /api/v1/data-access/azure/container-contents - name: run_wdl_workflow description: Submit a WDL workflow to the Cromwell server. source_endpoint: POST /api/v1/cromwell governance_note: 'Any real deployment of these tools should treat job submission, abort, move and delete operations as write/consequential actions requiring explicit human approval, and must respect the workspace airlock: CloudOS is the control plane for a Trusted Research Environment, and participant-level data must not be surfaced to an agent context.' deployment: mode: none verified: derived tools: 20 checked: '2026-08-12' source: catalog MCP census