jupiter-lend
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation references the installation of official SDK packages
@jup-ag/lendand@jup-ag/lend-read. These packages are part of the protocol's official development toolkit and are managed by the project maintainers. - [INDIRECT_PROMPT_INJECTION]: The skill involves fetching and processing data from the Solana blockchain, which constitutes a potential surface for indirect prompt injection from untrusted metadata.
- Ingestion points: Data is ingested through SDK methods
client.vault.getAllVaults()andclient.vault.getAllUserPositions()as described inSKILL.md. - Boundary markers: The instructions lack specific boundary markers or instructions for the agent to treat blockchain-retrieved data as untrusted or to ignore embedded instructions.
- Capability inventory: The skill provides functionality for transaction signing and network interaction with the Solana blockchain via the protocol SDKs, allowing for financial operations.
- Sanitization: There are no explicit sanitization or validation steps described for blockchain metadata before it is processed by the AI agent.
Audit Metadata