pCloudy-mcp-tool

Documentation

Documentation link · transport stdio

Tools

• get_all_application — Get list of all the application
• upload_app — This will upload the application to the user account
• install_app — This will install the app on the devices
• resign_app — This will re-sign the app to make it available for installation on iOS devices
• get_browser_list — Get list of available browser from pCloudy for specified machines
• book_browser — Book a browser on a specified VM
• get_browser_live_view_url — Get the live view URL for a booked browser session
• release_browser — Release a booked browser on a specified VM
• get_device_list — Get list of available devices from pCloudy for specified platform
• book_device — Book a device from pCloudy using device name (e.g.,
• get_device_details — Get detailed information about a specific device by ID
• get_live_view_url — Get live view URL for a booked device
• release_device — Release a booked device using booking ID (RID)
• refresh_authentication — It refresh the user authentication
• get_qpilot_projects — Get list of QPilot owned projects
• create_qpilot_project — This will create a new project for Qpilot run
• get_qpilot_folders — Get list of QPilot owned folders
• create_qpilot_folder — This will create a new folder for Qpilot run.
• get_qpilot_tests — Get list of QPilot tests and their test steps details, organized by project and folder
• create_qpilot_test — Create a new test in QPilot. Requires folderId to specify which folder the test will be created in. (*mandatory) Requires test name
• delete_qpilot_project — Delete a project in QPilot. Requires project name to specify which project to delete. (*mandatory)
• delete_qpilot_folder — Delete a folder in QPilot. Requires folder name to specify which folder to delete. (*mandatory)
• delete_qpilot_test — Delete a test in QPilot. Requires test name to specify which test to delete. (*mandatory)
• create_qpilot_playground_session — Create a QPilot playground session for a specific test. Requires test id to specify which test to create a session for. You can get the test list using the get_qpilot_tests tool. (*mandatory) Requires platform information to know which platform to run the session on. All platforms (*mandatory) For mobile platforms (android/ios), requires device id to know which device to run the session on. You can get the device list using the get_device_list tool. (*mandatory for mobile) For web platform, requires vm_id, browser name(only chrome allowed) and browser version to know which browser to run the session on. You can get these from the get_browser_list tool. (*mandatory for web)
• close_qpilot_playground_session — Close an active QPilot playground session for a specific test. Requires test_id to identify which session to close.
• add_qpilot_step — Add and execute a single natural language step in an active QPilot playground session. Requires test_id of the active session (*mandatory) and step_content, the natural language description of the action to perform (*mandatory). A playground session must already be open via
• insert_qpilot_step — Insert a new step at a specific position in an active QPilot playground session. Requires test_id (*mandatory), insert_at (0-based index where the step will be inserted, *mandatory), and step_content (natural language description, *mandatory). Existing steps at and after insert_at are shifted forward. A playground session must already be open via
• delete_qpilot_step — Delete a step at a specific position in an active QPilot playground session. Requires test_id (*mandatory) and step_idx (0-based index of the step to delete, *mandatory). A playground session must already be open via
• restart_qpilot_test — Restart execution of all steps from the beginning in an active QPilot playground session. Requires test_id (*mandatory). A playground session must already be open via
• get_qpilot_test_steps — Get the list of steps for a specific QPilot test, including step ID, index, content, and execution status. Requires test_id (*mandatory). Useful before retrying or deleting a specific step.
• get_qpilot_test_data — Get the test data (key-value pairs) stored for a specific QPilot test. Requires test_id (*mandatory).
• add_qpilot_test_data — Add or update test data key-value pairs for a QPilot test. Existing keys are preserved; new keys are added and duplicate keys are overwritten. Requires test_id (*mandatory) and test_data as a JSON object string e.g. \'{
• delete_qpilot_test_data — Delete specific keys from the test data of a QPilot test. At least one key must remain after deletion. Requires test_id (*mandatory) and keys as a JSON array string e.g. \'[
• modify_qpilot_test_data — Modify the values of existing keys in the test data of a QPilot test. Only keys that already exist can be modified; use add_qpilot_test_data to add new keys. Requires test_id (*mandatory) and test_data as a JSON object string with the keys to update e.g. \'{
• retry_qpilot_step — Retry a single step in an active QPilot playground session. Requires test_id (*mandatory) and step_idx (0-based index of the step to retry, *mandatory). Only that specific step will be re-executed. Use get_qpilot_test_steps to see current steps and their indices. A playground session must already be open via
• retry_qpilot_steps_from — Retry all steps from a specific step onwards (including that step) in an active QPilot playground session. Requires test_id (*mandatory) and step_idx (0-based index of the step to start retrying from, *mandatory). The given step and all steps after it will be re-executed. Use get_qpilot_test_steps to see current steps and their indices. A playground session must already be open via

About MCP

The Model Context Protocol (MCP) is an open protocol Anthropic introduced for connecting LLM-based agents to external tools and data sources. Providers publish MCP servers that expose their API surface as structured, discoverable tools — an MCP-compatible client (Claude Desktop, Cursor, Cline, Continue, etc.) can connect to the server and call its tools without any per-provider integration code.

Browse every MCP server on the APIs.io network or compare with the broader Agent Skill surfaces of the same providers.