relay

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

• Third-party content exposure detected (high risk: 0.90). The skill makes live REST calls to the public Relay API (e.g., POST https://api.relay.link/quote and GET https://api.relay.link/intents/status) and the bridge workflow (see src/commands/bridge.rs and SKILL.md) directly parses steps[].items[].data.to and .data from the API and uses them to construct and submit onchainos contract calls, so untrusted third-party responses can directly influence executed actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill performs runtime calls to https://api.relay.link to fetch quote "steps" (steps[].items[].data) that provide the contract "to" address, calldata, and value which are then passed to onchainos wallet contract-call and executed, so remote content directly controls executed transactions.

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

• Direct money access detected (high risk: 1.00). The skill is explicitly designed to move cryptocurrency. It provides a "bridge" command that executes cross-chain asset transfers, fetches quotes with execution steps (to, data, value), and invokes an onchainos wallet contract-call to submit on-chain transactions. It returns transaction hashes/request IDs and includes wallet login checks and broadcasting instructions. These are specific crypto/blockchain transaction capabilities (signing/broadcasting transfers), so this is direct financial execution authority.