A powerful orchestration skill for chaining AI agent tasks into sequential, multi-step processing pipelines.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install pilot-task-chain
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install pilot-task-chain using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The pilot-task-chain skill provides a robust framework for building sequential workflows where the output of one AI agent becomes the input for the next. This capability allows developers to decompose complex, monolithic tasks into a series of smaller, manageable steps handled by specialized agents. By leveraging the pilot-protocol and the pilotctl command-line utility, it bridges the gap between individual agent actions and sophisticated end-to-end automation.
At its core, this skill is designed for high-interdependency environments within the Openclaw Skills ecosystem. It excels at handling data transformation, multi-stage research, and complex deployment scenarios where data must flow seamlessly through a predefined logical path. By utilizing this skill, users can ensure that intermediate results are accurately routed and processed, providing a level of control and reliability that single-agent systems often lack.
To get started with this skill, you must have the pilot-protocol installed and the pilotctl binary present on your system PATH. The pilot daemon must be active.
# Start the pilot daemon if not already running
pilotctl daemon start
# Ensure jq is installed for result parsing
# On macOS/Linux:
sudo apt-get install jq # or brew install jq
Configure your environment variables for the agents you intend to use in your pipeline (e.g., AGENT_1, AGENT_2).
The skill relies on JSON-formatted task metadata managed by the pilot-protocol. The following table describes the primary data structure used during task chaining:
| Attribute | Type | Description |
|---|---|---|
| task_id | String | The unique identifier for a specific step in the chain. |
| status | String | Current state (submitted, running, completed, or failed). |
| result | String/JSON | The actual data output generated by the agent. |
| agent_id | String | The identifier of the agent performing the specific task step. |
Loading
A professional smart agriculture assistant for yellowhorn farming that analyzes multi-depth soil sensor data to provide real-time irrigation decisions.

A decentralized coordination skill for AI agents to join, discover, and establish secure mesh trust within named swarms.

A strategic planning tool for architecting website hierarchies, prioritizing page development, and mapping growth-focused URL structures.

A comprehensive strategic framework for planning SEO roadmaps, prioritizing tasks, and navigating the AI-driven search landscape.

A specialized AI skill designed to maximize recurring revenue and customer retention through Amazon Subscribe & Save program optimization.

A powerful task orchestration skill designed to distribute independent workloads across multiple AI agents simultaneously and aggregate their results.








































