A decentralized governance framework on Base Sepolia that allows AI agents to follow rules, log actions, and prove compliance on-chain.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-constitution
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 agent-constitution 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 AgentConstitution Skill?
AgentConstitution provides a robust framework for regulating AI agent behavior through smart contracts on the Base Sepolia network. By integrating this skill into the Openclaw Skills ecosystem, developers can ensure their agents operate under a public constitution, which is essential for earning user trust through verifiable on-chain actions. This skill bridges the gap between autonomous execution and human-defined ethics by providing a "legal" layer for AI.
The primary purpose of this skill is to allow agents to interact with a suite of governance contracts, including registries, action logs, and tribunals. This ensures that every significant action is transparent and every agent remains accountable to a set of genesis rules, such as preserving human override and preventing self-modification of governance logic.
AgentConstitution Skill Use Cases
- Verifying if an AI agent is compliant before initiating high-value transactions or delegating sensitive tasks.
- Programmatically retrieving active constitutional rules to guide agent decision-making processes.
- Logging significant agent actions on-chain to provide a tamper-proof audit trail for users and regulators.
- Monitoring global emergency status via the KillSwitch contract to pause operations during network or protocol instability.
How AgentConstitution Skill Works
- The agent or developer queries the AgentRegistry contract to verify the compliance status of a specific agent ID.
- The skill fetches active constitutional rules from the Constitution contract to determine the boundaries of allowed behavior.
- Before performing a transaction, the agent checks the KillSwitch contract to ensure no global emergency is active.
- During execution, the agent logs significant actions (Transaction, Delegation, etc.) to the ActionLog contract with associated risk levels.
- In the event of a rule violation, the Tribunal system processes slashing penalties based on the predefined genesis rules.
AgentConstitution Skill Setup
To get started with this skill on Base Sepolia, use the following configuration:
# Set your RPC URL and Private Key
export RPC_URL=https://sepolia.base.org
export AGENT_PRIVATE_KEY=your_private_key
# Check if an Agent is currently compliant
./scripts/check-compliance.sh <agentId>
# Read all active constitutional rules
./scripts/get-rules.sh
AgentConstitution Skill Data Schema & Taxonomy
The AgentConstitution skill manages data across several specialized contracts to maintain the integrity of the Openclaw Skills environment:
| Data Component | Format | Description |
|---|---|---|
| Agent Registry | Mapping | Maps agent IDs to compliance status and registration metadata. |
| Action Log | Event/Struct | Contains actionType, riskLevel, contextHash, and a text description. |
| Constitution | Array | A list of active rules, including genesis rules and human-proposed amendments. |
| Governance State | Boolean/Uint | Tracks emergency status and the threshold for rule activation. |
AgentConstitution Skill Advanced Features
- Human-in-the-loop governance allowing non-agents to propose and endorse new rules using USDC staking.
- Multi-level risk categorization for logged actions, ranging from Low to Critical.
- Immutable Genesis Rules that prevent agents from causing harm or self-modifying their governance structure.
- Automatic slashing mechanisms that penalize non-compliant agents based on rule severity.
SKILL.md
Loading
Related Openclaw Skills
GemBox Skill
A comprehensive developer tool for building .NET applications that process Excel, Word, PDF, and email files using GemBox components.
PDF Watermark Tool
A smart PDF utility that automatically detects page orientation to apply perfectly centered, adjustable text watermarks.
Content Moderation Skill
A two-layer safety framework designed to protect AI agents from prompt injections and sensitive content violations.
Model Router Manager
An intelligent multi-model routing manager that optimizes AI performance and reduces costs through automated model selection and failover logic.
Feishu Edge TTS Voice Sender (Windows)
A specialized Windows skill for converting text to speech and sending it as native Feishu voice bubbles.
memory-qdrant
A local semantic memory plugin for Openclaw Skills providing vector-based context storage and recall without external API dependencies.
Featured*
