defillama-research

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

• Third-party content exposure detected (low risk: 0.10). The skill queries the DefiLlama public API (api.llama.fi, yields.llama.fi, stablecoins.llama.fi), which returns third-party protocol-authored names, descriptions, and metadata, constituting low-risk indirect prompt injection via fetched web API data.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill references an external JSON-RPC/HTTP mutation endpoint retrieved from a local configuration file during runtime execution, representing a weak operational dependency.