Safe Multisig Skill

Propose, confirm, and execute Safe multisig transactions using the Safe{Core} SDK (protocol-kit v6 / api-kit v4). TypeScript strict. Use when an agent needs to operate a Safe smart account — (1) create/predict a new Safe, (2) fetch Safe owners/threshold/nonce, (3) list pending multisig txs, (4) build + propose a tx, (5) add confirmations, (6) execute a tx onchain, or (7) troubleshoot Safe nonce/signature issues across chains (Base/Ethereum/Optimism/Arbitrum/Polygon/etc.).

This skill appears to implement legitimate Safe multisig functionality, but the package metadata understates the runtime requirements. Notable points to consider before installing: - The scripts require SAFE_SIGNER_PRIVATE_KEY (env var) to sign/deploy/propose/approve/execute transactions — providing that key grants the ability to sign transactions that can move funds. Do not provide your main private key. - The skill also uses RPC endpoints (RPC_URL / --rpc-url) and interacts with the Safe Transaction Service (api.safe.global). You should verify or override those endpoints to ones you trust. - The registry entry claims no required env vars; that is incorrect. Treat the skill as requiring a signing credential and an RPC/API configuration. - If you proceed: run the code in an isolated environment, review scripts (especially propose/execute/approve/create-safe), and prefer a low-privilege or ephemeral signer (or a hardware/custodial signing flow) with spending limits. Consider running the scripts in read-only mode first (safe-info, list-pending) and only provide a signer after manual code review. - If you are not comfortable auditing the TypeScript, do not supply private keys to this skill. If you must automate, use a signing service or HSM that enforces policy rather than raw private keys in environment variables.

Type: OpenClaw Skill Name: safe-multisig-skill Version: 2.1.0 The OpenClaw skill is designed to manage Safe multisig wallets, which inherently involves sensitive operations like using private keys for signing and interacting with blockchain networks. The code demonstrates strong security practices, including robust input validation for addresses and transaction hashes, restricting file access to the workspace (`scripts/propose-tx.ts`), and explicitly handling private keys via environment variables without any evidence of exfiltration. The `SKILL.md` and `README.md` files provide clear instructions aligned with the skill's purpose and do not contain any prompt injection attempts or instructions for malicious behavior. Several 'FIX' comments in the code indicate a proactive approach to addressing potential vulnerabilities, such as preventing TOFU attacks (`scripts/execute-tx.ts`) and path traversal (`scripts/propose-tx.ts`). All network calls are directed to legitimate Safe Transaction Services (e.g., `api.safe.global`) or specified RPC URLs.