Host Shared Task Queue

    2

    Async task delegation for AI agents via shared folders—perfect for cross-OS and remote worker coordination.

    $9

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    host-shared-task-queue

    Example session with this skill installed

    Run the video render for project_v1 on the Windows host using the render-queue and tell me when it's finished.

    • Read your context and instructions
    • Compiled the host-shared-task-queue

    UUID: 8f2b-4e1a...
    Waiting for result from 'render-queue'...
    Done!

    {
  "status": "success",
  "output_file": "C:\\Renders\\project_v1.mp4",
  "duration_seconds": 145,
  "artifacts": ["log.txt", "thumb.png"]
}

    Connects securely to your tools. The creator never sees your data.

    About this skill

    What it does

    The Shared Task Queue skill provides a robust, low-tech way for an AI agent to delegate tasks to external processes using a shared filesystem directory as the transport. It implements an asynchronous inbox/outbox pattern that works across different machines, operating systems, and user accounts without requiring complex network setup or message brokers.

    Why use this skill

    Developers often face the "walled garden" problem where an AI agent in a Linux container needs to trigger a process on a Windows host, a legacy machine, or a browser session behind a NAT. This skill bridges that gap. Unlike standard MCP servers or HTTP webhooks, it requires zero infrastructure beyond a shared folder (SMB, NFS, Dropbox, or SyncThing). It handles the messy details of atomic file writes, cross-OS line endings, and polling logic.

    Supported tools

    • Compatible with any watcher process that can read and write JSON files.
    • Built-in bash scripts for send, wait, receive, and purge operations.
    • Fallback support for Python 3 when specialized CLI utilities are missing.
    • Supports shared directories via Samba, SyncThing, NFS, or local mounts.

    The Result

    The agent submits a structured JSON payload and receives a structured response once the external worker completes the task. All communication is persisted in the filesystem, providing a clear audit log and making failures easy to debug.

    • Not for high-frequency streams due to filesystem polling latency.
    • Requires pre-configured shared folder permissions.
    • No built-in task retries; responsibility lies with the external worker.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 6 months ago

    Software
    Pythonboth ends to speak MCP

    Permissions

    Terminal / Shell
    Network Access
    Listed6 months ago

    What's inside

    Frequently Asked Questions

    This skill is ideal for AI agents running in restricted environments (like Linux containers) that need to control tasks on a host machine, legacy hardware, or remote systems without setting up complex webhooks or message brokers.

    The skill is environment-agnostic; it works with any agent capable of executing shell commands or Python. It is designed to bridge operations between different OS environments, such as a Linux agent communicating with a Windows-based worker.

    No specialized infrastructure is required. You only need a shared directory accessible by both the agent and the worker process, such as a local mount, an SMB/NFS share, or a cloud-sync folder like Dropbox.

    The package includes the core skill logic, a set of ready-to-use Bash scripts for task management (send, wait, receive, purge), and Python fallback scripts for environments where Bash is unavailable.

    The skill uses a JSON-based inbox/outbox pattern with atomic file operations to prevent data corruption and race conditions, ensuring reliable task delegation even over slower network mounts.

    New in Agents & Orchestration