Security audit
Security checks for vulnerabilities and agentic risk
This skill is not malicious, but it should be reviewed carefully because it asks users to configure a private key and documents irreversible token-changing actions without enough safety scoping.
Install only if you understand the contract and wallet being used. Use a dedicated low-balance key or secure signer, verify every recipient/spender/amount/contract address, avoid broad approvals, and treat mint/burn as privileged irreversible operations that may not apply to standard ERC20 tokens.
The setup instructions require configuring a PRIVATE_KEY but provide no guidance on secure secret handling, which can lead users to paste raw keys into unsafe environments, logs, prompts, or shared configuration files. Because the skill can perform on-chain write actions, compromise of that key could directly enable theft of funds, malicious approvals, or privileged contract interactions.
The skill documents multiple write functions that can transfer tokens, set allowances, burn assets, or mint tokens, but it does not include an explicit warning that blockchain transactions are irreversible and may spend funds or grant dangerous approvals. In an agent-assisted workflow, this omission makes accidental loss, overbroad approvals, or unintended state changes more likely because users may treat these operations as ordinary reversible API calls.
The skill is presented as a generic ERC20 interaction tool, but it exposes mint and burn operations that are not part of the standard end-user ERC20 surface and can materially change token supply. In an agent context, documenting these state-changing methods without prominently scoping them to privileged/admin contracts increases the risk of accidental or unauthorized supply manipulation attempts, especially if users assume all listed functions are routine token operations.
No suspicious patterns detected.
