xiond-usage

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The skill's scripts and documentation (e.g., SKILL.md, references/network-config.md, and scripts/query-*.sh such as query-tx.sh and query-balance.sh) call public RPC endpoints (e.g., https://rpc.xion-testnet-2.burnt.com:443) and allow overriding node-url, ingesting untrusted on-chain/user-generated JSON (txs, memos, logs) which the agent parses and uses to determine status and subsequent actions, so third-party content could indirectly influence behavior.

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

• Direct money access detected (high risk: 1.00). The skill exposes explicit cryptocurrency wallet and transaction functions: it supports mnemonic-based account creation/restoration, local key management, and a concrete "send-tokens.sh " command that constructs, signs, and submits token transfer transactions (returning txhash). These are direct blockchain financial execution capabilities (wallets + sending transactions), not generic tooling. Therefore it grants direct financial execution authority.