Parallel MCP server

Read from the package source without running it. The installed server may list more.

• add_runs_to_task_group

  Initiates multiple task runs within a TaskGroup, allowing batch execution of tasks with specified inputs and output requirements.

• create_findall_runwrite action

  Initiates a FindAll run to search for entities matching specified criteria. Returns immediately with a queued run object; use the returned run ID to poll status, stream events, or receive webhook notifications as the search progresses.

• create_findall_spec_from_objectivewrite action

  Converts a natural language search objective into a structured FindAll specification that can be used to execute targeted searches. The generated spec serves as a customizable starting point and can be further refined by the user.

• create_monitorwrite action

  Create a web monitor that periodically executes a search query at a specified frequency and sends updates to an optional webhook. The monitor runs immediately upon creation and then continues according to the configured schedule.

• create_task_groupwrite action

  Creates a new TaskGroup to organize and monitor multiple task runs together. Use this to establish a logical grouping for related tasks that should be tracked as a unit.

• create_task_runwrite action
• delete_monitorwrite action

  Permanently delete a monitor and stop all its future executions. Deleted monitors cannot be updated or retrieved.

• extract_content_from_urls

  Extracts relevant content from specified web URLs, optionally focused on a search objective or keyword queries.

• get_event_group

  Retrieve a specific event group associated with a monitor. The response contains event items that represent individual events within the group.

• get_monitor

  Retrieve a specific monitor by its ID. Returns the complete monitor configuration including status, cadence, input settings, and webhook configuration.

• get_task_group_runwrite action

  Retrieves the status and details of a specific task group run by its run ID. The run result data is available separately via the `/result` endpoint.

• get_task_runwrite action

  Retrieve the status and details of a specific task run by its ID. Use the `/result` endpoint to access the run's output results.

• get_task_run_inputwrite action

  Retrieves the input data that was provided to a specific task run. Use this to inspect what parameters or data were passed when the task run was executed.

• get_task_run_resultwrite action

  Retrieves the result of a completed task run by its ID. This operation blocks until the task run finishes executing, then returns the final result.

• get_taskgroup

  Retrieves a TaskGroup and its aggregated status across all runs, providing a summary view of the group's execution state.

• list_monitor_events

  Retrieve events for a specific monitor, including errors and material changes from up to the last 300 event groups. Events are returned in reverse chronological order with the most recent first.

• list_monitors

  Retrieve all active monitors for the user with their current configuration and status. Supports cursor-based pagination to efficiently browse large monitor lists.

• list_monitors_paginated

  Retrieve a paginated list of active monitors ordered by creation time, newest first. Use cursor-based pagination to navigate through results.

• list_task_group_runs

  Retrieves all task runs within a TaskGroup as a resumable stream, with optional inclusion of run inputs and outputs.

• search_web

  Search the web for information based on natural language objectives or keyword queries. Returns relevant results with excerpts optimized for the specified use case.

• simulate_monitor_event

  Simulate sending an event to a monitor to test its event handling and triggering behavior. Useful for validating monitor configurations without waiting for real events.

• stream_task_group_events

  Streams real-time events from a TaskGroup, including status updates and run completions. The connection remains open for up to one hour while at least one run in the group is active.

• stream_task_run_eventswrite action

  Stream events for a specific task run, including progress updates and state changes. Event frequency is reduced for task runs created without event streaming enabled.

• update_monitorwrite action

  Update an existing monitor's configuration, including its search query, check frequency, webhook URL, and associated metadata. At least one field must be provided to apply changes.

• Code scan5 source files scanned25/25
• –Live reliabilityno gateway calls yet and no remote to proben/a
• –Tool poisoningtools not inspected (local package is not executed); not countedn/a
• Auth qualitylocal package, no credentials required12/15
• Maintenancelast push 138 days ago8/15
• Maintainer identityregistry namespace matches repository owner7/10

Safety score

80/100

higher than 55% of 866 Scraping and browsing servers

Tools

24

the typical Scraping and browsing server has 6

Tools that change things

50%

typical: 0%. Fewer write tools means less an agent can break

GitHub stars

31

more than 87% of 583 Scraping and browsing servers

Downloads a week

152

more than 65% of 309 Scraping and browsing servers with a package

Parallel: common questions

Is Parallel MCP server safe?

Mostly: it is graded B (80/100). Read the Parallel safety report

How do I install Parallel?

It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.

Does Parallel need an API key?

Not as far as the registry entry and our scan can tell: no credentials are declared or required.

Is Parallel maintained?

The last commit was 176 days ago (2026-05-12). The latest release is v1.0.2.

What can I use instead of Parallel?

Servers from other publishers that do the same job: Brightdata MCP server, Browser Use MCP server and Pinkpixel Dev Web Scout MCP server. Compare all Parallel alternatives.

How does Parallel compare with other Scraping and browsing MCP servers?

safety score 80/100 (higher than 55% of 866 Scraping and browsing servers); tools 24 (the typical Scraping and browsing server has 6); tools that change things 50% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.