sui-client

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt includes an explicit CLI example that requires inserting the user's 12-word recovery phrase verbatim into a command (sui keytool import '<12-WORD-PHRASE>'), which encourages the agent to request and echo a highly sensitive secret into its output — a high-risk secret-exfiltration pattern.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to query the external MCP server https://sui.mcp.kapa.ai at runtime for up-to-date answers and verification, so content fetched from that URL can directly influence agent prompts/outputs.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly describes wallet and transaction functionality on the Sui blockchain: creation and storage of keypairs and a 12‑word recovery phrase, instructions to import keys (sui keytool import), CLI commands for transaction operations (explicitly instructing to use "sui client ptb" for transaction operations, merge coins, etc.), commands to check balances and gas, and SDK/faucet functions (requestSuiFromFaucetV2). These are specific crypto/blockchain capabilities that enable signing and submitting token transactions (i.e., moving funds). Therefore it grants direct financial execution authority.