elfa-api

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md and elfa-api-reference.md explicitly instruct the agent to make live calls to https://api.elfa.ai (e.g., /v2/data/top-mentions, /v2/data/keyword-mentions, /v2/data/trending-narratives and /v2/chat) which ingest and return user-generated public content from Twitter/X and Telegram that the agent is expected to parse and act on (summaries, narratives, chat analysis), so untrusted third‑party content can materially influence its outputs.

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

• Direct money access detected (high risk: 1.00). The skill explicitly includes a built-in pay-per-request payment flow (x402) that requires on-chain USDC payments on Base. It documents getting a 402 payment response, signing a USDC transfer authorization, resending with an X-PAYMENT header, and even provides code that uses a private key / EVM signer (privateKeyToAccount, toClientEvmSigner) and the @x402/fetch wrapper that "handles payment automatically." These are specific crypto/Blockchain payment controls (wallet signing and settlement on-chain), not generic API calls — i.e., the skill is explicitly designed to authorize and send crypto payments for requests.