A comprehensive TypeScript toolkit for AI agents to manage Safe multisig account lifecycles, from creation to transaction execution.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install safe-multisig-skill
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 safe-multisig-skill 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 Safe Multisig Skill?
The Safe Multisig Skill is a professional-grade suite of TypeScript scripts designed to enable AI agents to interact with Safe smart accounts. Built on the Safe{Core} SDK (protocol-kit v6 and api-kit v4) and ethers v6, this skill allows for secure, programmatic management of multisig wallets across various EVM-compatible chains. By integrating this into the Openclaw Skills framework, developers can give their agents the ability to navigate complex governance and treasury tasks.
This skill focuses on technical accuracy and security, providing a bridge between high-level agent logic and low-level blockchain interactions. It handles the nuances of the Safe Transaction Service, ensuring that multisig transactions are properly proposed, indexed, and executed with full validation of owners, thresholds, and nonces.
Safe Multisig Skill Use Cases
- Deploying new Safe accounts with specific owner sets and confirmation thresholds.
- Automating the proposal of treasury transactions for DAO or team management.
- Collecting off-chain signatures from multiple owners to fulfill multisig requirements.
- Executing batch or individual transactions once the required signature threshold is reached.
- Auditing pending transaction queues and troubleshooting nonce synchronization issues.
How Safe Multisig Skill Works
- The agent initializes the environment by bootstrapping dependencies and configuring network-specific RPC URLs.
- When a transaction is needed, the agent generates a structured JSON request following the internal schema.
- The skill uses the Safe API Kit to fetch the current state of the Safe, such as the owner list and the next valid nonce.
- A transaction hash is computed and signed off-chain using the agent's private key.
- The signed transaction is proposed to the Safe Transaction Service for other owners to see and approve.
- Once the threshold of signatures is met, the execution script submits the bundled signatures to the blockchain via an RPC provider.
Safe Multisig Skill Setup
To get started with this skill within the Openclaw Skills ecosystem, navigate to the skill directory and run the bootstrap script to install dependencies and prepare the TypeScript environment.
cd safe-multisig-skill
./scripts/bootstrap.sh
# Perform a sanity check to verify network connectivity
./scripts/safe_about.sh --chain base
Ensure that you have set the SAFE_SIGNER_PRIVATE_KEY and appropriate RPC_URL in your environment variables before attempting to sign or execute transactions.
Safe Multisig Skill Data Schema & Taxonomy
The skill manages data through strict JSON schemas to ensure compatibility between the AI agent and the Safe protocol.
| Component | Format | Description |
|---|---|---|
| Transaction Request | JSON | Contains to, value, data, and operation as defined in tx_request.schema.json. |
| Network Mapping | Markdown | tx_service_slugs.md maps chain slugs like base or arbitrum to their API endpoints. |
| Script Output | JSON | Standardized output including safeTxHash, nonce, and confirmation status. |
| Config | Env Vars | Manages sensitive data like private keys and API keys securely. |
Safe Multisig Skill Advanced Features
- Support for deterministic Safe deployment (predicting addresses before they are on-chain).
- Multi-chain compatibility across Ethereum, Base, Polygon, Optimism, and Arbitrum.
- Detailed error reporting and input validation using TypeScript-strict patterns.
- Capability to handle off-chain signature collection via the Safe Transaction Service API.
- Flexible configuration allowing for custom transaction service URL overrides for private networks.
SKILL.md
Loading
Related Openclaw Skills
OP0 Altar Protocol Skill
A comprehensive automation tool for AI agents to deploy and manage self-rewarding meme coins on the Solana blockchain through the OP0 Altar protocol.
Osori
Osori is a technical local project registry and context loader designed to help AI agents and developers manage repository metadata and environment switching.
Agent Avengers
A comprehensive multi-agent orchestration skill that decomposes complex tasks into subtasks for specialized, parallelized AI agents.
docx-to-md
A robust automation tool to convert .docx files into structured Markdown while preserving formatting and extracting images.
Safe Multisig Skill
A TypeScript-powered toolkit for AI agents to manage Safe smart accounts and execute multisig transactions across multiple EVM chains.
x402janus-acp
A specialized agent skill that enables hiring the x402janus security scanner via the Virtuals Agent Commerce Protocol (ACP) for automated wallet auditing.