ton-proof

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill requires taking a user- or service-provided payload verbatim and passing/returning it (included in the generated proof JSON and in the call to generate_ton_proof), which forces the LLM to handle and output arbitrary secret values if the payload contains them.

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

• Direct money access detected (high risk: 1.00). The skill explicitly provides wallet/crypto signing functionality for TON: it generates a TonConnect proof (signed payload) via the generate_ton_proof tool and returns wallet address/publicKey/signature, and notes it requires signing access. This is a specific crypto wallet signing capability (listed in the policy as crypto/blockchain wallets/signing), so it must be flagged as Direct Financial Execution risk.