AgentMesh Task

Run any AgentMesh task with dynamic JSON parameters

AgentMesh Task

AI-generated

Summary

Run an AgentMesh task on a specified platform with configurable JSON parameters, optionally targeting a specific online client or allowing automatic client selection by AgentMesh.

Inputs

  • Client — Select a specific online client to run the task on, or leave empty to let AgentMesh auto-select the least busy client.
  • Platform (required) — Choose the target platform for the task; determines available task types.
  • Task Type (required) — Choose the task type for the selected platform, which defines the kind of AgentMesh task to execute.
  • Parameters (required) — Enter task parameters in JSON format. Defaults to an automatically loaded template based on the selected platform and task type, or empty if none is available.

Output shape

a list of task result items each as a JSON object with task details and metadata

For each input item, the node posts a task to AgentMesh API, polls the task status until success or failure (or timeout), then outputs the task results. If the API returns multiple result items under common keys, each item is output separately with metadata; otherwise, the raw result object is output. On failure or timeout, the node throws an error unless 'continue on fail' is enabled, in which case the error message is output.

Examples

Example 1: Run a keyword search task on a platform

Set platform to the desired one; select task type (e.g., 'keyword search'); edit parameters as needed in JSON; optionally select a client or leave for auto-selection.

Links

Discussion