solana-trader

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

• Third-party content exposure detected (high risk: 0.90). The skill's monitor.js and scan.js explicitly fetch and parse public, user-generated token and pool data from third‑party APIs (e.g., https://api.dexscreener.com/... and https://api.geckoterminal.com/...), and the agent directly reads and interprets that untrusted content to score opportunities and drive trading actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill at runtime fetches transaction payloads and execute endpoints from Jupiter (https://lite-api.jup.ag), deserializes remote transaction bytes, signs them locally and submits them for on-chain execution, so this URL supplies remote-executed instructions and is a required runtime dependency.

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

• Direct money access detected (high risk: 1.00). The skill explicitly provides cryptocurrency financial execution capabilities: it includes a solana_wallet tool that creates and stores keypairs (wallet creation/address/balance), a solana_swap tool that "executes a token swap via Jupiter aggregator," and an automated trading monitor (cron script) that uses solana_scan to find opportunities and solana_swap to execute trades. Environment configuration (position size, stop loss/take profit, max positions) and state files (encrypted wallet keypair, trades/state) further indicate automated on-chain transaction execution. These are specific crypto/blockchain tools for moving funds, so it meets the Direct Financial Execution criteria.