A specialized skill for exchanging structured datasets with automated schema negotiation and metadata tracking over the Pilot Protocol.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install pilot-dataset
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install pilot-dataset using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is pilot-dataset?
The pilot-dataset skill provides a robust framework for agents to exchange structured data such as CSV, JSON, and Parquet files. Unlike simple file sharing, this skill prioritizes schema negotiation and metadata enrichment, ensuring that both the sender and receiver agree on data structures and formats before the transfer occurs. By integrating with the broader ecosystem of Openclaw Skills, it allows for seamless data discovery and provenance tracking across distributed agent networks.
This skill is essential for complex workflows where data consistency and lineage are critical. It enables agents to publish dataset availability, request specific formats, and validate schemas programmatically, significantly reducing errors in multi-agent data pipelines and collaborative research environments.
pilot-dataset Use Cases
- Sharing CSV, JSON, or Parquet datasets with explicit schema definitions between agents.
- Negotiating data formats and transformations before initiating a large file transfer.
- Maintaining comprehensive dataset lineage and provenance metadata for auditability.
- Automated data discovery where agents broadcast available datasets to a peer network.
How pilot-dataset Works
- The publishing agent broadcasts dataset availability to peers using pilotctl, including format and row count metadata.
- A receiving agent sends a specific dataset request, optionally specifying a preferred format for the data.
- The publisher sends dataset metadata, including column definitions and schema details, to prepare the receiver.
- The actual data file is transferred via the secure file-sending mechanism of the underlying protocol.
- The receiving agent validates the incoming file against the previously negotiated schema to ensure data integrity.
pilot-dataset Setup
To use pilot-dataset, you must have the pilotctl binary installed and the pilot-protocol skill active. Ensure the daemon is running:
pilotctl daemon start
You will also need jq for parsing JSON responses and optionally python3 for automated format conversions. Ensure that pilotctl is correctly configured on your PATH to work with Openclaw Skills.
pilot-dataset Data Schema & Taxonomy
The skill utilizes structured JSON objects for negotiation and metadata tracking. The data organization follows this taxonomy:
| Component | Description |
|---|---|
| Dataset Availability | JSON containing name, format (csv/json/parquet), and estimated row count. |
| Dataset Request | JSON specifying the target dataset name and the preferred delivery format. |
| Dataset Metadata | Detailed schema mapping including column names and data types. |
| Provenance | Metadata identifying the source agent, version, and transfer timestamp. |
pilot-dataset Advanced Features
- Multi-format support including CSV, JSON, and Parquet for flexible data handling.
- Integrated schema validation logic to prevent processing of malformed data.
- Support for custom metadata fields to track domain-specific provenance.
- Seamless integration with other Openclaw Skills for complex multi-agent automation.
- Peer-to-peer dataset discovery via localized broadcast messages.
SKILL.md
Loading
Related Openclaw Skills
Pilot Discord Bridge
A bidirectional integration bridge that enables Pilot Protocol agents to send notifications to Discord and process incoming messages as event streams.
Windows Mouse and Keyboard Control
A PowerShell-based utility for AI agents to programmatically control mouse movements, clicks, and keyboard inputs on Windows systems.
Agentforce Persona Designer
A comprehensive framework for designing, scoring, and encoding high-fidelity AI agent personas for Salesforce Agentforce.
pilot-discover
A high-performance discovery tool for locating and filtering peers within the Pilot Protocol network based on capabilities and reliability.
pilot-directory
Maintain a persistent local directory and metadata cache for frequently contacted AI agents.
FlowZap Diagrams
A professional diagramming skill that converts natural language, code, or logs into valid FlowZap Code DSL with instant shareable playground URLs.
Featured*
