A decentralized peer review system providing independent verification and multi-party approval for AI agent task results.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install pilot-review
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-review 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 Review?
The Pilot Review skill establishes a robust quality control layer for AI agent ecosystems. It enables developers to implement a structured verification process where task results are reviewed by independent peers before final acceptance. This skill is built on the pilot-protocol and provides the necessary tools to orchestrate decentralized human or agent-in-the-loop validation, ensuring high-fidelity outputs in complex automation chains. By integrating this into Openclaw Skills, teams can enforce high standards and mitigate risks associated with autonomous execution.
Pilot Review Use Cases
- Quality control for sensitive task results before final system acceptance.
- Requiring independent verification from trusted peer reviewers for security audits.
- Implementing multi-party approval gates for critical infrastructure or financial outputs.
- Managing decentralized workflows that require manual or high-level agent oversight.
How Pilot Review Works
- The initiator completes a task or identifies an output requiring validation.
- Reviewers are selected based on criteria, such as tags or addresses, using the pilotctl command-line tool.
- Review requests containing the task ID and deadlines are broadcast to the selected peer addresses.
- Reviewers analyze the task data and submit approval or rejection messages via the pilot-protocol.
- The system monitors the inbox for incoming reviews and calculates the approval count against a defined threshold.
- Once the minimum approval count is met, the task is finalized and marked as approved.
Pilot Review Setup
To utilize this within Openclaw Skills, ensure you have the pilot-protocol skill installed and the pilotctl binary available in your system PATH. Start the background daemon using:
pilotctl daemon start
You will also need jq installed for processing JSON responses in the workflow scripts.
Pilot Review Data Schema & Taxonomy
The skill utilizes JSON-formatted messages for communication between peers. Below is the primary metadata structure used for reviews:
| Field | Description | Type |
|---|---|---|
| type | The message purpose (e.g., review-request, review-submission) | String |
| task_id | Unique identifier for the task being reviewed | String |
| deadline | ISO-8601 timestamp for review expiration | String |
| decision | The reviewer's verdict (approve or reject) | String |
| score | Numerical quality score ranging from 0.0 to 1.0 | Float |
Pilot Review Advanced Features
- Customizable approval thresholds to require consensus or simple majority logic.
- Automated reviewer selection based on specialized metadata tags like senior-auditor.
- Integration with the pilot-protocol inbox for asynchronous result processing across distributed networks.
- Support for multi-agent validation chains within Openclaw Skills environments.
SKILL.md
Loading
Related Openclaw Skills
Pilot S3 Bridge
Bridges the Pilot Protocol with cloud storage services like AWS S3 and GCS, allowing AI agents to manage files through secure gateway agents.
Pilot Reputation
Advanced reputation analytics and trend tracking for Pilot Protocol agent scoring.
Bubble Plot Generator
An automated visualization tool for generating gene expression bubble plots categorized by cell type and disease state.
pilot-role-assign
A sophisticated role orchestration skill for agent swarms that enables dynamic, capability-based assignment and hierarchical coordination.
Time Calc
A high-precision temporal toolkit that uses native shell commands to perform complex date and time calculations without external dependencies.
Pilot Relay
A store-and-forward messaging system for AI agents that ensures guaranteed message delivery even when peers are offline.
Featured*
